Recent draft: Developing a Laboratory Based Device to Test the Impact of Windborne Debris in Severe Wind Events Bachelor of Engineering (Civil and Project Management) Honours Project Report

Developing a Laboratory Based Device to Test the Impact of Windborne Debris in Severe Wind Events
Bachelor of Engineering (Civil and Project Management) Honours Project Report

[read more=”Click here to Read More” less=”Read Less”]
Executive Summary
The following report discusses the design of a testing apparatus and method for use that has been developed to perform windborne debris impact testing. This will enable the effects of windborne debris on building materials in the built environment to be determined and analysed
In order to develop a testing method in conjunction with the amendments in Australian Standard AS1170.2; Clause 2.5.7 (2011), extensive research and investigation in different testing methods and designs for an apparatus was required. Three different testing designs were investigated in order to determine the appropriate design to be developed. The most feasible and practical testing design chosen was a Pneumatic Cannon. There are many safety and instrumentation issues with such a testing apparatus, and they were investigated accordingly.
Once the design for the testing apparatus was established, the apparatus was constructed and then testing was performed. The apparatus constructed was a prototype rather than a full scale model due to time constraints and a change in the scope to the project. The results from the prototype testing method were quite promising.
The results of the prototype testing were recorded and analysed, the results show that there was a linear relationship between pressure and velocity. As the pressure increased in the testing system, the velocity of the projectile also increased. During testing, a nylon wad was placed inside the barrel behind the projectile. This aided in gaining an understanding of efficiency of the system and friction losses. The results concerning the wad were inconclusive pending further testing.
The findings were positive and the effectiveness of the final testing design is proven. It is therefore apparent that the prototype design can be adapted for a full scale version for impact testing.
Recommendations are provided for further testing to improve the accuracy of results, and for the development of a full scale windborne debris testing apparatus.
Throughout the project there was a lot of research required in order to develop the necessary knowledge to carry out the required tasks and achieve a result. The group wishes to formally acknowledge the contributions made to the project as it would not have been possible to carry out this work without the assistance given.
Firstly, the University of South Australia technical staff have played a pivotal role in both designing and constructing the testing apparatus and associated devices. Although the design team were ultimately responsible for producing the device, a large amount of advice was provided as well as carrying out the majority of the construction works.
The project supervisor, Professor Julie Mills, assisted the team throughout the course of the project by providing support and guidance when required as well as technical advice. Julie has also reviewed drafts of all of the documents produced and provided recommendations to improve these.
Dr. Bruce Wedding, a senior physics lecturer for the University of South Australia, provided assistance with the calculation of the required pressures to produce the specified velocities. This assistance was greatly appreciated and discussion with Bruce was very helpful throughout the course of the project.
Jason Sutton of SMC pneumatics aided the team in developing an understanding of the operation of pneumatic systems. Jason was also able to help a great deal with providing the names of potential suppliers and suggestions for certain components of the device.
The Marksman Firing Range of Adelaide gave members of the project team the opportunity to view one of their firing ranges and take note of the safety systems in place. This tour was very helpful in designing the safety system for the testing device and was greatly appreciated.
Finally, the firearms branch of the South Australian Police Force were very helpful in providing information relating to the legality of the testing device. The information provided was vital in ensuring that the project was legal and safe.
Overall, these contributions to the project have been most helpful and very much appreciated by the project team.
1.0 Introduction
1.1 Statement of Problem
The purpose of this report is to discuss the testing method that has been developed to perform windborne debris impact testing. The purpose of testing is to ensure building materials comply with the relevant Australian Standards.
In order to develop a laboratory based device to test for the impact of windborne debris, an extensive amount of research and investigation was required. Included in this research was different testing methods and designs for a testing device as well as safety precautions and many other topics.
Focusing on the design of the testing device, the most feasible and effective option for testing has been developed. The testing device that has been developed and outlined in the following report has been designed to test the impact of windborne debris on building materials in accordance with the requirements of Australian Standard AS1170.2; Clause 2.5.7(2011).
Included within the report is a literature review, research and construction methodology, construction outline of a prototype testing apparatus, testing results and a discussion of the lessons learned from the project.
A prototype apparatus was designed due to the extensive nature of the project. The prototype will provide the essential information to develop the testing method on a larger scale in the future.
There were three different testing methods identified throughout the group’s discussion and research. The advantages and disadvantages of each method are discussed in the following report as well as the reasons for selecting the final method.
The testing method was developed to test materials used in the building industry e.g. walls, roof sheeting, windows etc. The Australian standard required the apparatus to perform testing using a timber member of 4kg mass with a nominal cross section of 100mm x 50mm and a spherical stainless steel ball of 8mm diameter.
Paramount in achieving an effective design was ensuring that projectiles can reach necessary velocities to simulate wind speeds in different areas in Australia. The velocities required for testing from different wind regions in Australia have been calculated from AS 1170.2 and outlined in the report. The apparatus has been designed to have the ability to accelerate the specified projectiles to these velocities for testing.
1.2 Changes to AS1170
Due to the danger that windborne debris poses there was recently an amendment made to the Australian Standard AS/NZ 1170.2.
Previously Australian Standards AS/NZ 1170.2: 1989, Clause 3.4.7 stated,
In Cyclonic regions, windows shall be considered as potential dominant openings, unless capable of resisting impact by a 4kg piece of timber of 100mm x 50 mm cross-section, striking them at an angle at a speed of 15m/s. (Standards Australia, 1989 p.28).
The previous clause concerns dominant openings and only requires impact resistance for windows. It is known from previous research that building components such as Garage doors and Weather Board walls can be subject to damage from windborne debris. This is what has led to the update of the clause.
The updated clause requires the speed of impact for the testing to be significantly higher than the previous edition. The impact is also defined for both horizontal and vertical directions.
Australian Standards AS/NZ 1170.2: 2011, Clause 2.5.7 states:
Where windborne debris loading is specified, the debris impact shall be equivalent to-
a) Timber member 4kg mass with a nominal cross-section of 100mm x 50mm impacting end on at 0.4VR for horizontal trajectories and 0.1VR for vertical trajectories; and
b) Spherical steel ball 8mm diameter (approximately 2 grams mass) impacting at 0.4VR for horizontal trajectories and 0.3VR for vertical trajectories
Where VR is the regional wind speed given in Clause 3.2 (Standards Australia, 2011 p.13).
This amendment to the 2011 Australian Standard AS/NZ 1170.2 has led to an increase in the required impact strength of building materials. Building materials such as windows, wall panels and roof sheeting will need to satisfy the above clause to achieve certification
Developing a method for testing will provide confirmation for companies in industry of their products compliance with the amended clause. Compliance with the Australian Standards is of high importance for material manufacturers. Currently there is no device in Australia available to complete testing at higher wind speeds specified by the amended clause.
1.3 Research Questions
In order to provide direction for this project to be carried out of the following research questions were developed:
 Why is the impact of windborne debris important in wind events?
 What is the required velocity for testing?
 How was the testing carried out prior to the amendments?
 How do other countries complete their testing for impact of windborne debris?
 How can the velocity of the testing projectile be measured?
 What OHS&W issues arise from a project of this nature?
 What is the most feasible testing apparatus?
 What are the legal requirements for this type of impact testing?
 Are there any operating issues with the design during testing?
2.0 Literature Review
Due to the damage caused by strong wind events around the world, testing the ability of building products to withstand the forces that these events exert on them is of high importance. In particular, the damage caused by wind borne debris is tested very extensively, especially in the United States of America where tornadoes and hurricanes are common.
In order for the project team to gain some understanding of the importance of the testing as well as some ways in which it may be carried out, this literature review has been undertaken. This review will consider the following areas: severe wind events and wind borne debris damage, the design and testing standards relating to wind borne debris in Australia and other countries around the world, and the types of testing that have been carried out in the past including the methodology used.
2.1 Severe Wind Events and Windborne Debris
Tropical Cyclones and tornadoes are severe weather events that rapidly move over large sections of land and possess the ability to cause a large amount of destruction. Houses, large buildings and structural frames are at risk of being damaged by storm surges and wind gusts. Debris from existing buildings or surrounding areas can become loose due to wind gust uplift and then be propelled vast distances at high speeds. These free materials and parts of structures can effectively become missiles flying through the air and cause damage to surrounding houses, buildings and people.
Boughton et al. (2011) defines wind as:
Air movement relative to the earth, driven by several different forces, especially pressure differences in the atmosphere, which are themselves produced by differential solar heating of different parts of the earth’s surface, and forces generated by the rotation of the earth. The differences in solar radiation between the poles and the equator, produce temperature and pressure differences. These, together with the effects of the earth’s rotation set up large-scale circulation systems in the atmosphere, with both horizontal and vertical orientations. The result of these circulations is that the prevailing wind directions in the tropics, and near the poles, tend to be easterly. Westerly winds dominate in the temperate latitudes
For all types of severe wind events the wind is highly turbulent or gusty. This is produced by eddies or vortices within the air flow which are generated by frictional interaction at ground level, or by shearing action between air moving in opposite directions at altitude.
2.1.1 What is the Impact of Wind Events and Cyclones?
Cyclones and strong wind events impose a great danger to cities, suburbs and human life. They do not only impact single structures but entire villages and cities. Tropical cyclones, also popularly known as hurricanes or typhoons, are among the most spectacular and deadly geophysical phenomena. Both the most lethal and the most expensive natural disasters in the United States of America’s history were tropical cyclones. Emanuel (2008) explains that the Galveston Hurricane of 1900 killed nearly 8000 people, and Hurricane Andrew of 1992 did more than $35 billion in damage. Globally, tropical cyclones rank with floods as the most lethal geophysical hazards. For example a single storm in Bangladesh in 1970 killed nearly half a million people (Emanuel, 2008 p.75).
Other effects associated with wind events can also cause damage. For example De Scalley (2008) investigated the frequency and consequences of cyclones in the Cook Islands and noted that the Southern Cook Islands are more than twice as frequently affected as the Northern Islands. However, past cyclone disasters in the latter group suggest that risk to human life is greater there due to the potential for inundation of the atolls by storm surges. He also explains that half of cyclones with human impacts have occurred during El Niño events, with weak to moderate El Niños almost as important in this respect as strong El Niños. Only 13% of cyclone impacts have occurred during La Niña events.
In spite of their importance and the publicity they receive, the physics of tropical cyclones are not well known. They are natural disasters that are extremely expensive and, once over, require a large amount of cleaning up and rebuilding. People can lose their livelihood, homes and life. The event of such severe wind events impacts on individuals, communities and countries along with the larger scale global economy.
2.1.2 What is the Impact of Windborne Debris in Wind Events and Cyclones?
Windborne debris created by strong wind events, particularly hurricanes and cyclones, is a major source of damage to the built environment. Holmes (2011) explains that large wind gusts and cyclones cause damage to buildings due to the increased wind load. They compete with seismic loading as the dominant environmental loading for an effected structure. Strong gusts of wind and cyclones can cause an equal amount of damage as an earthquake over a long period of time.
During a strong wind event the envelope of a building must remain intact in order to prevent internal pressurization. If not then the internal and external pressures can exceed the levels that the structure has been designed for previously. This can lead to an increase in failure of buildings and a subsequent increase in the injection of additional debris into the wind stream. The cumulative damage imposed by debris may lead to breaching of the building envelope and hence cause further damage to the interior contents and overall structure of buildings.
Extensive studies of building performance have concluded that wind-borne debris is a principal cause of building envelope breaches during a strong wind event (Grayson et al., 2012). The most common wind-related damage associated with light-frame wood construction (such as houses etc.) was roof sheathing panel failures, which was a significant contributor to the total damage to buildings during Hurricanes Andrew, Charley, Ivan, and Katrina (Grayson et al., 2012).
From the information above it is clear that windborne debris can endanger human life and damage infrastructure in cities and towns. Windborne debris can also be represented mathematically. When debris strikes a building, damage that will be incurred is proportional to the debris kinetic energy. For example an expression for damage caused by a single piece of windborne debris can be represented as:
?= 12???3?2= 12???3?2?2 (Wills et al., 2002).
This equation uses a constant of proportionality of unity. To combine this equation with flight conditions for a solid three-dimensional object and eliminate the typical missile dimension the equation will be expressed as:
?= 116??{(????)(????)}3?2?8 (Wills et al., 2002).
??= Mass per cubic metre of material
?? = generalized force coefficient
?? = Air density
??=the ratio of forces between friction coefficient of unity for a loose object and the wind force of a fixed object required to break it loose.
J= Kinetic energy (this is specific for the type of material and speed at impact)
U= Wind Speed
This equation can provide assistance in designing a testing mechanism that will measure the extent of damage a material can withstand upon impact.
2.2 Projectile Materials and Standards
2.2.1 Projectile Materials
In a real wind event, any object can become a projectile. However, standardised testing methods for this phenomenon tend to specify typical projectile objects to be used in laboratory simulation. Debris impact testing and research undertaken by Texas Tech University (TTU 2003), using techniques similar to those being considered for our project, used debris that included a piece of timber with a weight of 6-8kg and cross-section of 38 x 89mm. This compares with the 4kg, 50 x 100mm piece of hardwood which will be used throughout our project as specified by AS 1170.2, Clause 2.5.7 (Standards Australia 2011). Also projected by TTU were a 34kg, 76mm diameter steel pipe and a variety of PVC pipe. Debris such as heavy steel pipe mentioned above is not particularly useful for the research and testing we are conducting as such a material is not considered common in an Australian/New Zealand cyclone. Although materials such as sign posts similar to the steel tested may become windborne in a cyclone it is not as common and as practicable to test compared with smaller more common debris such as a ball bearing or piece of hardwood. The 8 mm ball bearing is specified as representative of the missiles that can develop when areas of gravel road or footpath surfaces become airborne in cyclones and high wind events.
2.2.2 Wind Speeds
Kempler (2010) of NASA’s Goddard Earth Sciences Data and Information Services Centre found maximum surface wind speeds of up to 77.8 ms-1 occurred during the devastating Hurricane Katrina in August of 2005, which is classed in the worst hurricane category. When maximum wind speeds from devastating Australian cyclones are compared to those of hurricane Katrina, there are considerable differences. For example anemometer readings recorded throughout cyclone Yasi that occurred in Queensland in early 2011, show maximum 10 minute mean wind speeds ranging from 17 to 38 ms-1 and maximum 3-second gust speeds ranging from 25 to 52 ms-1 (Boughton 2011). The study established 67ms-1 as a likely overall upper limit of gust speed anywhere affected by Cyclone Yasi, however this has not been confirmed and is only an estimate through extensive research. During another devastating Australian Cyclone, Cyclone Tracey, maximum wind speeds of up to 60.23ms-1
were measured before measuring instrumentation was destroyed (The Australian Bureau of Meteorology, 1974). Through wind speeds obtained of two of Australia’s worst cyclones in Australia it is apparent that they do not produce the same maximum wind speeds as a country like America. This is most likely the reason as to why we see heavier debris being used for impact testing in places such as the Texas Technical University as speeds experienced in American hurricanes have the ability to carry heavier materials.
Although the speed at which materials will be launched will vary depending on the local wind category, speeds ranging between 10 and 50ms-1 will include the worst case situations required in Australia. In comparison the above mentioned materials projected at TTU were undertaken at higher speeds, varying from 34ms-1 to 76ms-1. This is mostly due to the higher velocity of winds needed to simulate tornadoes and hurricanes in America.
2.2.3 Standards
Comparing the American wind codes with the AS/NZS codes it is apparent that there are similarities between projectile materials used for debris impact testing. Recent American model and building codes describe projectile material used for debris impact testing as most commonly a 2×4 inch (50 x 100 mm) projectile. Also used as a projectile on a less common basis is a small steel ball, replicating gravel which may become airborne in hurricanes and tornadoes (Crandell et. al, 2002). ASTM E (1996) discusses the typical large projectile used in wind debris testing as a piece of lumber with a weight of 9-lb (4 kg) and dimensions of 2×4 inch (50 x 100 mm) impacting the testing product end on with a speed of 34 mph. This is equivalent to 15.2 ms-1, which is in the range of the worst case impact testing velocities outlined in the AS/NZS wind code. From these dimensions it seems likely that the Australian code has adopted the requirements of the US code in terms of the mass and dimensions of the timber projectile required for testing.
The testing which our project group will undertake on windborne debris will be in accordance with AS/NZS 1170.2, Clause 2.5.7 where the debris impact shall be equivalent to:
a) Timber member 4kg mass with a nominal cross-section of 100mm x 50mm impacting end on at 0.4VR for horizontal trajectories and 0.1VR for vertical trajectories; and
b) Spherical steel ball 8mm diameter (approximately 2 grams mass) impacting at 0.4VR for horizontal trajectories and 0.3VR for vertical trajectories
where VR is the regional wind speed given in Clause 3.2, (Standards Australia, 2011).
Table 3.1 in the AS/NZS Regional Wind Speeds code provides formulae to calculate VR in cyclonic regions within Australia. Once determined the debris impact can be calculated for each material mentioned above. Although the regional wind speed will vary for what different companies wish to test depending on whether they sell their product in cyclonic areas or not, our team will simulate the worst case regional wind speed for Region D of 99 x FD, where FD is constant of 1.1. Therefore, a worst case regional wind speed in a cyclonic area will result in debris impact velocities ranging from 10.9ms-1 vertical to 43.6ms-1 horizontal using the debris impact equations given in AS/NZS 1170 Clause 2.5.7.
Although the material which the debris will be launched upon is not specified in the code it will be beneficial to test impact of debris on materials which have caused houses and buildings to fail during cyclones. (Boughton, 2011) states that “the level of damage to roller
doors during cyclone Yasi was significantly greater than any other component of post 80’s housing.” Despite some of the analysed house being over 20 years old it is still common for houses to fail due to large dominant openings such as roller doors and large windows being damaged during cyclones. For this reason throughout our report we will be concentrating on debris impact testing materials which account for dominant openings in residential houses and buildings such as roller doors.
2.3 Testing Methods for Windborne Debris
Wind borne debris testing is achieved through the use of a testing apparatus, which projects objects at a target. The target is tested for the amount of force which it can withstand from the projected objects, being a function of the object’s velocity and mass. It is also found that the velocity at which the testing apparatus projects the debris is dependant not on the size of the object, but on the mass per unit area, the product of density and thickness (Wills et al. 2002). There are forces including aerodynamics, trajectory, and damage due to velocity and resultant forces which can be presented in the form of mathematical equations. These are shown to accurately predict the velocity of different projected objects (Wills et al. 2002). Calculated velocity predictions were compared to measured velocities in a testing facility at the Colorado State University, the Wind Engineering and Fluids Laboratory (WEFL). This laboratory has two wind tunnels which are used in studying the effects of wind on buildings and structures, in particular ultimate wind loads and debris impact testing (Colorado State University 2008).
2.3.1 Types of Testing Equipment and Methods
The testing conducted at the WEFL is highly relevant to our debris testing project as it describes the behaviour of testing 2 x 4 timbers as well as spherical steel balls of different diameters, which are the same cross-sections as those described in the Australian Standard AS 1170.2. Testing at the WEFL facility was conducted using three large boundary layer wind tunnels which produce wind speeds using large fans to analyse trajectories and flight paths (Wills et al. 2002).
Wind tunnels could be an option for wind borne debris impact testing, however in this form of testing rig it would be very difficult to control the flight path and rotations of the projectile, which is particularly important when testing with 2 x 4 planks. The WEFL wind tunnel facility for testing was used to analyse the flight paths and trajectories of the 2 x 4 planks. This includes the plank’s rotation and bouncing off of the floor before impacting the target (Wills et al. 2008). This is not an ideal form of testing for our wind borne debris impact testing rig as we want to control the direction and orientation of the plank so that it impacts the object being tested straight and perpendicular. This is important to be controlled as it will cause the most damage and is thus the worst case scenario for a collision. Therefore using a large wind tunnel which allows the 2 x 4 planks to fly without restrictions or control was not appropriate for our testing rig.
The WEFL facility also has a Missile Impact Facility (MIF) which is specifically used to test the ability of building materials to resist impacts from objects in high velocity wind situations. It is apparent that the MIF has specifications and applications which are very similar to what we require for wind borne debris impact testing. The MIF produces a high speed release of any missile type projectile, including 2 x 4’s, or 50 mm x 100 mm cross section planks (CSU 2008). The MIF uses a nitrogen gas powered cannon with a release valve, propelling the
missile down a confining tube to keep it straight and on target. The velocity of the missile upon release is measured using an electro-mechanical set up before it covers the 6 ft. (1.83 metres approx.) gap from release to impact into the material being tested (CSU 2008). The missile described as being used in the MIF is a 12ft. (3.66 metres approx.) long 2 x 4 plank, and is projected at speeds of up to 110 mph (176 km/h or 48.9 ms-1) (CSU 2008). As noted previously, the maximum required impact speed required for testing to the Australian Standard is 43.6 ms-1 (157 km/h). Therefore it is apparent that this speed is achievable through the use of a MIF type rig.
The Australian code AS 1170.2 requires a 2 x 4 (50mm x 100mm) plank which weighs 4kg, meaning that the plank needs to be relatively long, depending on the type of wood and thus its density. Therefore the MIF shows that it is possible for a very long plank to be projected (3.66 metres) and accurately used as a testing projectile. It also shows that this type of testing rig is ideal for testing using long and heavy projectiles, rather than fan forced wind tunnels which do not control the behaviour of the plank in the air including rotations and direction.
There are other examples of testing rigs for debris impact similar to the MIF created by Universities and organisations, mostly in the United States of America. Another example of a testing apparatus for wind borne debris comes from the state of Texas, at the Texas Tech University (TTU), Wind Science & Engineering Research Centre. The Texas Tech University claim to be leaders in the field of wind borne debris impact testing, with hundreds of tests being conducted in their Debris Impact Test Facility (TTU 2003). The testing rig used at TTU is very similar to that of the MIF, using a cannon type propulsion to project a missile down a cylindrical containing tube before being released into a material to be tested. This rig uses a pneumatic cannon which accelerates a 6 – 8 kg, 2 x 4 to a speed of 67 m/s (241.2 km/h). Similarly to the MIF an electrical gate type system is used to record the velocity of the projectile upon release (TTU 2003). The main difference between this and the MIF is that it uses air compression for propulsion, whereas the MIF uses nitrogen compression for propulsion. This confirms the suitability of this type of testing rig for the applications of projecting 2 x 4 planks of 4 kg, and steel objects of smaller dimensions.
The testing rig at TTU is also used to test the damage caused by debris falling vertically after being picked up by a tornado. Research conducted at TTU found that objects will reach speeds of up to 67 mph (29.8 m/s) (TTU 2012) and as such this speed is also tested. Impact testing in our rig will not be tornado specific, as it will be focussed on extreme wind speeds reached in cyclonic events. However this shows that vertical speeds from falling debris will generally be lower than that of the horizontal extreme wind event, meaning that vertical speed testing is easily achievable.
Another issue that must be addressed by any testing methodology and apparatus is what areas of the material panels being tested are most vulnerable to damage. Hence the position to be aimed for as a worst case scenario must be considered as part of a defined testing procedure. In the procedure for testing a panel, it must withstand 3 missile projectiles according to TTU (2012). The first missile is aimed at the most vulnerable area, generally in the middle of the panel, and the remaining two missiles are aimed at connections or supports of the panels. There are definitions of pass and fail criteria for the panels which are specific to American standards and as such can only be used as a general guide. We must establish our own pass and fail criteria according to Australian Standards.
The materials used to construct the testing rig at TTU are detailed as follows:
 Air Tank – 30 gallon
 Electric Over Air Valve
 4 inch Aluminium Coupler to connect the barrel to the valve
 4 inch x 20ft. long PVC barrel
 Optical Timing Sensors
 Signal Conditioners
 Two Precision Timers
 Control Panel for pressure, aiming, and firing
 Horizontal Articulation Cannon Carriage with controls
 Hydraulic Scissor Lift for adjusting vertical height
 Steel frame for supporting the rig (TTU 2012)
The materials used in our testing rig will ultimately be different, but this can be used as a guide for what might be required to create a wind borne debris impact testing cannon.
2.4 Safety
After researching and considering the type of apparatus that is to be constructed and commissioned throughout the course of this project it has become clear that the task is extremely hazardous. With this in mind occupational health, safety and welfare (OHS&W) had to be considered a matter of the utmost importance.
2.4.1 OHS&W Policies
The OHS&W policy of The University of South Australia (2012) states the aim of achieving ‘zero harm’ and making the work place safer for all involved. It is stated within this policy that all appropriate measures must be taken in order to ensure that all tasks are carried out in the safest possible manner.
With regard to the device being built, it is clear from this OHS&W policy that a great many safety measures will need to be put in place in order to pass the University’s safety specifications.
During the construction of this device all appropriate safety regulations must be adhered to in the workshop. Upon completion, the device will need to be used in order for a Standard Operating Procedure to be developed. SOP’s ‘are written instructions for tasks that outline the preferred and safest method of performing that task in a standardised manner.’ (UniSA 2011). The operating procedure will need to be checked and approved by the appropriate persons as stated in the University’s OHS&W policy. Throughout the course of the lifetime of the device this procedure will need to be reviewed and updated to ensure the university’s goal of continuous improvement is realised.
2.4.2 Risk Management Strategies
There have been many safety issues discovered through this literature review. These include: projectiles rebounding upon impact with the test specimen, projectiles penetrating the test specimen and the test specimen or the projectile breaking apart causing shrapnel to become a risk. In order to minimise the risk caused by these events safety screens will be required in order to create safe viewing areas. In addition to this, there will be noise issues during the operation of the device and as such hearing protection will be required. During
the development and construction of this apparatus many other safety issues will arise and as these are realised they will need to be dealt with accordingly.
All of this information is relevant to the project as it is another item that requires consideration and development throughout the design and commissioning of the apparatus.
2.5 Conclusion
In conclusion, this literature review has provided a great deal of insight into the dangers of wind borne debris as well as the testing strategies that have been employed throughout the world. The findings from this research have been beneficial to the completion of the project and aided in developing an effective and efficient testing device.
Through this research it has been found that wind borne debris can be the major cause of damage during significant wind events such as hurricanes, tornados and cyclones. It is clear that products’ resilience to wind borne projectiles is of the utmost importance as their failure often results in the failure of structures due to the creation of dominant openings that increase the wind loads on structures beyond their design values.
Although similarities exist between testing in Australia and overseas, differences have also been noticed. The major difference found relates to the size and velocity of the projectiles used for the testing. Also, it has been noted that the worst case wind speeds in other countries are significantly higher than those found in the larger storm events in Australia.
Many different testing devices have been found through research with major differences between them, some more economically viable than others.
The literature review has provided a reasonable understanding of the results that should be achieved as well as exposing a myriad of issues that require attention in order to succeed in producing an effective testing device.
3.0 Testing Equipment and Procedures
3.1 Propulsion Options
After reviewing the literature relating to windborne debris testing and consulting with a range of technical experts, three possible options for propulsion were established. Each option was investigated with pros and cons of each being discussed. Rather than detailing a design for the three possible testing mechanisms, each was assessed in terms of their feasibility. This enabled certain options to be eliminated as a propulsion mechanism before valuable time was spent on creating a detailed design. One of the three proposed options was later selected as the final testing mechanism that was then further developed and tested. A detailed design of the selected testing mechanism can be found later within the report.
The following three testing devices were investigated:
 Elastic Energy Apparatus
 Pitching Machine
 Pneumatic Cannon
Constant to all three possible testing devices was the need to satisfy the specifications detailed in the Australian Standard. In this case the required material was a piece of timber with cross section 100mm x 50mm weighing upwards of 4kg and an 8mm diameter stainless steel spherical ball of a mass 2 grams.
3.1.1 Elastic Energy Apparatus
Unlike other possible testing option’s that were selected mostly due to previous success in similar projects, the elastic energy apparatus was considered due to its simplicity. The apparatus is a fairly basic and easily understood concept as illustrated in Figure 3.1. The way in which the elastic apparatus would work involves the projectile being placed in a groove to guide the material. One end of the projectile is free while the other end is placed against an elastic material such as a large-scale rubber cord, similar to a bungee cord. The elastic material would then be pulled back to the required distance, allowing suitable elastic potential energy to build up. Once an appropriate elastic potential is available, releasing the elastic cord will fire the projectile.
Figure 3.1 – Elastic Energy Apparatus Concept
The elastic energy apparatus has both advantages and disadvantages. Fewer moving parts within the mechanism mean that the construction would be relatively straightforward and the design would be cost effective. The elastic material is the only material within the design that would require significant movement. The size of the apparatus however would need to be quite large to reach the required testing velocity. Accuracy of the projectile would also be problematic as the possible rotation of the projectile could be hard to control. While the projectile would be fired along a guide or rut, this would need to be enclosed to ensure that certain factors, such as lifting force, do not affect the projectile’s direction during firing. This will ultimately cause the projectile to hit the testing material at an angle that is not 90 degrees. The guide would also increase friction, reducing the acceleration and velocity of the projectile.
The complexity of measuring the potential energy needed to reach the required impact velocities would also create problems. While basic calculations can be made, a trial and error approach would need to be taken until the necessary impact velocities are reached. Accurate calculations would become very difficult due to unknown friction losses and not knowing how the projectile will perform in the air once fired. Not only would a trial and error approach take up a large amount of time, but it also requires the use of unnecessary resources and extra materials such as projectiles. Also disadvantageous, is the ongoing calibration requirements for the elastic material. Over time the elasticity of the material will change and decrease. This would need to be constantly monitored and the elastic regularly changed to ensure accuracy of the device is as high as possible.
While such a mechanism is possible to design and construct, it is not practical for the project due to its potential inaccuracy and the required maintenance needed once constructed.
Elastic Material
3.1.2 Pitching Machine
Following on from the proposed elastic energy apparatus was the discussion of a pitching machine to carry out testing (Figure 3.2). Pitching machines, like those used for baseball and cricket practice sessions are most commonly designed with two internal high velocity counter-rotating discs that enable balls to be projected at high speed. The first disadvantage which the pitching machine poses is that the velocities required for the testing cannot be achieved without major technical changes to a standard “off the shelf” pitching machine. Manipulation in order to achieve higher velocities would require the addition of a more powerful motor in order to increase the turning speed of the discs. Research has shown that an exceptionally powerful motor, capable of producing large amounts of torque, would be required to increase make the testing velocities achievable. The use of such a motor poses a number of issues such as the size of the device as well as the high purchase and operating costs.
Figure 3.2 – Pitching Machine Concept
Further difficulties involved with designing the windborne debris impact testing apparatus through use of a pitching machine included adjusting the rotating discs. The disc spacing would need to be increased and decreased in order to accommodate the 100mm x 50mm sized piece of hardwood and the small ball bearing. This will require considerable alterations to the pitching machine, firstly to accommodate such a large motor but also to allow the large piece of wood to be fed through the counter rotating discs easily.
The long list of disadvantages involved with potentially using a pitching machine design for testing far outweighs any advantages. It can be concluded that the pitching machine is unrealistic and not a feasible means of conducting the testing required for the project.
3.1.3 Pneumatic Cannon
The literature review determined that a pneumatic cannon has been successfully utilised for windborne debris testing in the United States. In general it was found that the pneumatic cannon has been made up of the following – a pressure vessel, a valve capable of achieving a high flow rate and a circular barrel.
In order to operate the pneumatic cannon there must be a source of compressed air available. This is achieved with either an air compressor or bottled compressed air depending on the pressure required. The operation of pneumatic cannons varies depending on the chosen design. The operation of the cannon that was built for this project is relatively similar to some of those found through the literature review.
The pressure vessel must be connected directly to the large diameter valve that provides pressure straight to the barrel upon opening. The pressure vessel is charged to the required air pressure for the velocity to be tested. The valve is then opened, at which point the air travels into the barrel and applies a force to the projectile causing it to accelerate and leave the barrel to collide with the test specimen.
Figure 3.3 – Pneumatic Cannon
Given the size of the projectile is specified to be a required dimension, the diameter of the barrel must be large enough to accommodate this. The actual size of the barrel will depend on the section sizes available. If steel was the chosen barrel material, it would need to be 125mm in diameter with a 10mm wall thickness. Based on early estimates, the optimum barrel length was decided to be 5m. This value however is, as mentioned, an estimate as any calculations do not consider friction and other influencing factors.
The pressure vessel is required to have 4 to 5 times the capacity of the barrel, as advised by a representative of SMC pneumatics in Adelaide. This extra capacity is to ensure that force is applied to the projectile for the entire length of the barrel. It is difficult to make theoretical estimates relating to required pressures for the desired velocities due to the large number of variables involved that could not be accurately defined, and as such trial and error had to be used to determine the pressures used for testing. Due to safety reasons it was decided that the pressure would not exceed 1000KPa and as such the velocity may have been limited depending on the efficiency of the device.
The valve used to transfer energy from the pressure vessel to the barrel could be one of a number of different options. One option considered was a 2 inch (50 mm) diameter actuated ball valve. Another option for consideration was a diaphragm valve of the same diameter. The diaphragm valve has the advantage of opening more rapidly however the flow rate it can provide is not as high as that of the ball valve. Both options were evaluated during construction, as discussed in the next chapter.
In addition to these components the device required a number of different fittings to connect the parts. A second valve was required to allow air to enter the pressure vessel and a pressure gauge was required in order to determine the charge contained within the pressure vessel.
3. 2 Instrumentation Options
3.2.1 Measuring Velocity – Different Considered Options:
An important aspect of the design method for developing an impact testing facility for windborne debris is the ability to accurately measure the speed of the projectiles at the point of impact. It is critical to accurately measure this speed as the standard AS/NZ 1170.2 states that the testing must occur with a projectile impact of a specific speed. Therefore it needs to be verified that the correct testing procedure has been achieved by accurately recording the speed of the projectile.
There are several different ways in which the speed of the projectiles can be measured, and they all have their own advantages and disadvantages for use in the projectile mechanism.
The considered possibilities were:
 A radar gun similar to that used by police to detect the speeds of cars
 A video camera which will be later reviewed in slow motion and projectile velocities calculated
 Hand-held stopwatch
 A series of light gates
 Proximity sensors
Radar guns used by police to detect speeding cars use an emitted radio frequency which rebounds off of an object, such as a car, and is then received back by the gun. Due to the Doppler Effect the returning radio wave has a longer frequency when the car is travelling away from the gun, and a shorter frequency when the car is travelling towards the gun. From this frequency difference the speed of the car can be calculated. This form of speed detection could be applied to the impact testing facility fairly easily and would give accurate and consistent results. However, the main difficulty with a radar gun is that it could not be built into the impact testing facility as it would require manual operation by a person. This may raise safety concerns as the person operating the radar gun would need to be close to the projectile and inside safety barriers. Readings would also be subject to operator error.
Research into specific radar guns found that for the project, a special gun would be required that was able to detect very small objects that are travelling very fast. Most radar guns used for speed detection of cars measure speed in 100 milliseconds, whereas the gun needed for this project measures speed in 10 milliseconds and speed can be recorded off-parallel. This means that a person can be using the radar gun from the side of the projectile, or behind or at any angle, which would help safety concerns as the safest position to record speed from could be established and utilised. Most radar guns have an “angle-error” if they are not used from directly behind the moving object. Another advantage of using a radar gun would be that speeds could be recorded as close to impact as possible, rather than recording speed at the barrel and assuming a loss over the flight path. It is therefore the most accurate option, however it was difficult and expensive to obtain an appropriate radar gun compared to other options, and as such it was decided that this option was not feasible.
16 has quoted the “Solo2 Stalker DIGITAL Speed Radar Gun” at a price of $2,750, which was decided to be too expensive.
A video camera used to record the projectile flight and impact would be analysed after the testing procedure to determine the time of flight over a specified distance, which would then give an average speed of the projectile. The camera would be positioned to view the flight path from the side, and there would be some form of incremental scale behind the projectile’s flight path so that the distance travelled over a time could be calculated. Using a video camera has the advantage of being used for the second purpose of analysing the actual impact. The impact of the projectile into the material being tested would be recorded and could be reviewed in slow motion, which would aid the companies using the impact testing facility as they could analyse how their product failed. However using this method has the disadvantages of not giving a result instantly, time needs to be taken to go back and review the video to calculate the projectile’s speed, rather than an instant number being given by say, a radar gun. It also has the disadvantage of giving an average speed over a defined flight path distance, rather than an instantaneous speed at impact. After consideration it was decided that this option was not feasible as it was tedious and not highly accurate. Video cameras were still used during the testing procedure to record results and for later analysing the impacts; however speeds were not calculated through this method.
Using a hand-held stopwatch was the simplest and cheapest method considered for recording the projectile’s speed. Similarly to the video camera method, the time of flight would be recorded and therefore speed calculated over a defined flight path distance. This method has the obvious disadvantage of being heavily dependent on human error. It was considered extremely difficult for a person to accurately press the stopwatch buttons between the start and impact of the projectile, especially as the entire flight time of the projectile was around a quarter to half a second. Therefore this method was not appropriate for the impact testing facility as it is not accurate enough.
3.2.2 Chosen Option
Light gates and proximity sensors work in a similar way, where there is a path between an emitter and a receiver, and when this path is broken a signal is recorded. To use light gates or proximity sensors there are two gates or sensors positioned in the flight path of the projectile at a specified distance apart, and when the projectile is fired the time difference between each of the two paths being broken will be recorded. This will be used with the known distance to calculate the speed of the projectile. This method was deemed to be most suitable for the testing device as it is accurate and gives instant results. Information on the specific design of this apparatus is given in chapter 4.
3.3 Safety Issues
The size and type of testing device developed resulted in many safety issues that had to be considered, testing projectiles at high speeds with compressed air has the potential to cause serious injury. During testing, operators must be a safe distance away and protected from any safety hazards, as must any instrumentation equipment or nearby structures.
The safety issues that had to be considered during this project included:
 Noise
 Rebounding of projectiles
 Projectiles going through a test specimen
 Containing air pressure
 Preventing the system from recoiling
Each of these will be discussed in the following sections.
3. 3.1 Noise
Noise-induced hearing loss is a common occupational injury in Australia. The two main factors that make noise in the workplace harmful are the loudness of the noise and the length of time a person is exposed to it each day. Noise destroys the delicate nerve cells in the inner ear that transmit sound messages to the brain.
Noise during testing can also affect concentration and fatigue of surrounding employees and students at the University. This can lead to lower productivity, stress and an increased risk of accidents.
3. 3.1.1 Noise Exposure
Part 2 of the 2010 Occupational Health Safety and Welfare regulations (Government of South Australia, 2010, pg. 23) explains the exposure standard for noise in industry is:
(a) an 8 hour equivalent continuous A-weighted sound pressure level, LAeq,8h of 85 dB(A) referenced to 20 micropascals; and
(b) a C-weighted peak sound pressure level, LC,peak of 140 dB(C) referenced to 20 micropascals.
A designer, manufacturer, supplier or importer of plant that may emit an unsafe level of noise must ensure that the plant is designed and constructed so that the noise emitted by the plant is, when installed and used in a reasonable foreseeable circumstance—
(a) so far as is reasonably practicable, not above the exposure standard; and
(b) to the extent that is reasonably practicable in the circumstances, as low as can be achieved,
A manufacturer, supplier or importer of plant that may emit an unsafe level of noise must make available to employers, on request, information about—
(a) the noise emitted by the plant; and
(b) ways to keep the noise to the lowest level that is reasonably practicable to achieve
The above standard explains that an unacceptable noise standard is an average exposure of 85 dB (measured on the A-weighted noise scale) over an eight-hour working day or a peak of 140 dB (for peak noise levels or exposure to varying, intermittent or impulse noise) (Government of South Australia 2010, pg. 23). If employees of the university and students performing testing are exposed to such noise, immediate damage to hearing is possible.
Based on this, the noise exposure levels were kept below this level during testing. The Australian Standards to not state an acceptable level of noise exposure in the work place
but simply provide an upper limit that must not be exceeded. It is stated that the lower the exposure to noise, the better and as such all exposure during testing was minimized to the lowest possible level. Noise Measures
Measures were taken to ensure noise exposure is minimized during testing. Minimizing the generation and emission of noise from the process of testing is a key priority in the Occupational, Health, Safety and Welfare Policy of The University of South Australia and legislation by the Government that is implemented by Safe Work SA.
Noise, when possible should either be completely eliminated, substituted, or a solution shall be engineered (Government of South Australia Safe Work SA, 2010).
The process of eliminating noise from testing the process involves completely removing the noise hazard or risk of exposure that is present. Removal of the hazard is the desired control solution. In certain situations/workplaces noise cannot be eliminated, this situation requires substituting or replacing components of testing/machinery in the workplace to minimize the effects of noise.
Most of the noise concerned with testing windborne debris cannot be controlled. There will be noise associated with releasing the pressure through the barrel as well as the noise created by the projectile impacting the test specimen. In addition to this there will be noise created by the running of the air compressor and other such things. Due to this, strategies were employed to reduce the effects of the noise created.
Some of these strategies include;
 Personal protective equipment (ear muffs)
 Remaining as far away from the device as possible during testing
 Remaining behind screens during testing
 Positioning signs to ensure that other parties remain a suitable distance away from testing
3. 3.2 Containing Projectiles
Projectiles can either rebound back off a test specimen or pass through a specimen. To contain projectiles, protective screening and dampening techniques needed to be used to ensure any ricochet or splinters from a projectile, or test specimen, did not injure an operator during testing.
The most appropriate option was to completely separate the operator from the test specimen and provide an opening only large enough for the projectile to enter the testing chamber. Clear screens were required to enable viewing of the impact damage and ensuring that the equipment was operating as designed.
Protection was also needed behind all test specimens. After an inspection of the Marksman Firing Range in Adelaide a number of options were discussed. One option is to use a pit of crumbed rubber (recycled car tyres) behind the projectile to contain it. During this inspection it was stated that any dense material that is able to absorb the energy of the projectile without being damaged is suitable for this purpose. An appropriate rear protection system was designed as discussed in the following chapter.
3. 3.3 Recoiling
Sir Isaac Newton’s third law states that for every action there is an equal and opposite reaction. That is, any force applied to an object will exert an equal force in the opposite direction. It is this reaction force, acting opposite to the applied force, which causes recoiling. If this force was not properly accounted for the cannon may move backwards upon firing presenting a hazard to operators as well as reducing the efficiency of the device.
Through research on the effects of recoiling on pneumatic cannons Associate Professor Brett Taylor at Radford University has established the recoil force applicable to firing a pneumatic cannon. Taylor (2006) used a compressed air cannon as a laboratory activity to investigate impulse, conservation of momentum, and kinematics. When the cannon was fired a force plate was placed under the cannon to measure the recoil force on the cannon. The impulse was then used to work out the final velocity of the projectile.
The experiment established a relationship between the firing velocity and force exerted on the cannon depending on the mass of the object. Taylor was able to further simplify the following form of the impulse-momentum relation.
F = mvf – mvi
F=Force on cannon (N)
vf = Final Velocity of projectile(at exit of barrel) (m/s)
vi = Initial velovity of projectile (m/s)
m = Mass of projectile (Kg)
As the projectile was at rest in the barrel the initial velocity was 0m/s; Therefore,
F = mvf
For Example if the projectile weighed 4kg and was fired at 40m/s
Force = 4 x 40
= 160N
Therefore, 160N of Force would be exerted on the pneumatic cannon at firing.
Any air resistance and friction of the wad in the barrel were ignored in this experiment. This is an indicator of the recoil force that may be present when firing such a projectile with pneumatic cannon. Such forces require systems to be put in place to counteract the recoil force.
Eric Kathe (2001) explains some design options that may apply to the recoil of a pneumatic cannon are;
 Double recoil system
 Active Recoil Suspension system
 Fixed System
The first two options that were investigated counteract recoiling force by acting as the
projectile is fired. This is carried out in the double recoil system by simultaneously firing masses in opposing directions thus cancelling the recoil force. The active recoil suspension system uses a spring mounting system to counteract the force. Both of these systems, whilst possibly very effective, were more complex than was needed to counteract the relatively small forces that were present during the cannon operation.
After discussion, it was found that the most suitable option for counteracting recoil force was to fix the cannon to an object with a very large mass.
Fixing the cannon to a large mass is both simple and cost effective as well as serving the dual purpose of placing the cannon off of the ground for operation. The system used is discussed later in the report.
3. 4 Legal Issues
After some discussion, it was decided that, given the nature of the project, it was necessary to discuss the legality of this testing device with the relevant authorities. A meeting was held with a representative from the South Australian Police at Holden Hill Police Station. There were concerns with whether the pneumatic cannon testing apparatus was considered as a weapon. Under the Firearms Act the cannon was suspected to be considered a compressed air gun during the meeting it was suggested that the firearms branch be contacted to further discuss the matter as there was no clear legislation relating to a device of this nature.
On this advice an email was sent to the firearms branch detailing the intentions of the project. A copy of this email can be seen in Appendix A. The aim of the email was to determine what is required, if anything, to operate this device legally and what design requirements may be in place to ensure it complies with the law.
After some time, the team was contacted with a response from the firearms branch. It was advised that, given the fact that the device is to be bolted down and not easily transported, it is classed as a cannon. Given this classification, it is not covered under the firearms act and as such it is not a requirement that permits be obtained. It was stated however, that there may be other laws brought into effect should any irresponsible or unsafe operation of this device lead to some sort of incident that causes harm or damage of any sort.
Based on this advice, the preliminary design of the device was deemed to be acceptable in regard to legal requirements.
3. 5 Wind Speeds
The regional wind speed (VR) of the area that the test specimen is to be constructed in is the deciding factor for the velocity at which testing will be carried out. Based on the regions defined in AS1170.2: Wind Actions, (see Figure 3.4 below) different wind regions will require vastly different velocities for testing. The regional wind speeds for all directions in testing are based on three second wind gust wind data.
Figure 3.4 – Australian Wind Regions (AS1170.2: Figure 3.1)
In Australia a large portion of the country is classified as being a non-cyclonic region (Regions A1-7, W, B). The coastal and near-coastal northern parts of the country e.g. top of Queensland, top of Northern Territory and North West of Western Australia, are rated in cyclone regions (regions C, D).
In the housing industry the design life of structures is usually considered to be 50 years. The average recurrence interval (R) is the inverse of the annual probability of exceedance of the wind speed e.g. for a 1 in 100 year wind event; R=1/100years. The average recurrence interval depends on the importance level of the structure. The collapsing of a house or erected structure can cause harm to people in the vicinity. The potential number of people who could be harmed leads to the design importance level. Structures, such as hospitals, are required for post disaster shelter and services. Buildings of this nature are designed to withstand a 1 in 2500 year event.
The design wind speeds in Table 3.1 have been calculated based on the following design lives and importance level for typical structures:
 For an importance level 2 (e.g. houses) – a 50year design life -1 in 500 year event
 For an importance level 3 (e.g. schools) – a 50year design life – a 1 in 1000 year event
 For level 4 post disaster structures (e.g. hospitals) – a 50 year design life – 1 in 2500 year event
Table 3.1 – Regional Wind Speeds for Each Wind Region
Wind Speed
Design Regional Wind Speed (m/s)
A (1-7)
These design regional wind speeds in Table 3.1 have then been modified for testing according to AS1170.2; Clause 2.5.7. The standard states that impact velocities will be tested in horizontal and vertical directions. For the horizontal direction the impact velocity shall be 0.4VR and the vertical direction impact velocity shall be 0.1 VR. The impact velocities for testing in each direction are shown in Table 3.2.
Table 3.2 – Design Wind Speeds for Each Direction of Testing
Horizontal Velocity
Impact Velocity (m/s)
A (1-7)
Vertical Velocity
Impact Velocity (m/s)
A (1-7)
As shown the maximum testing velocity required is 40m/s based on the regional wind speeds as specified by the Australian Standard. The highest testing velocities; 35.2m/s, 37.6m/s and 40m/s are applicable for testing for region D, which is the most severe cyclone region in Australia and occurs only on a small section of the Western Australian coast near Port Headland.
4.0 Construction
4.1 Construction Methodology
4.1.1 Selected Option
From the researched options discussed in Chapter 3 it was decided that the most feasible and practical system for the testing apparatus was a pneumatic cannon. This system was the most appropriate for the application of projecting long timbers and small steel balls. It was the easiest to calibrate for reaching different velocities simply by charging the pressure vessel to the required pressures to achieve the chosen velocity.
4.1.2 Developing a Prototype
Due to time constraints, safety concerns, high expense of required parts, and inexperience with the proposed systems it was decided to develop a smaller prototype. This prototype device enabled testing to ensure that the large scale version would be feasible within the space and budget.
Safety is of highest concern in any project, but is of particular importance when dealing with pressurized air and high velocity projectiles. As this testing apparatus had not been constructed before there were many unknowns. These included the ability of the pressure vessel to withstand high and changing pressures, the ability of the barrel to contain and transfer pressure to the projectile, and the behaviour of the projectile once it has left the barrel. Therefore the safest option is to conduct testing on a smaller scale, with lower pressures and a projectile which will cause minimal damage. A large tennis ball was chosen as it should have a predictable flight path, is soft and will cause minimal damage, and is a very good fit within the barrel. A prototype with low pressures and a soft projectile required more reasonable safety precautions, which were discussed in section 3.3.
Preliminary calculations relating to the pressure required to achieve certain velocities showed that a pressure of approximately 8 psi (55KPa) is required to achieve a speed of 40 m/s. This pressure is very low and, given that the equations ignored friction and other influencing factors, was deemed to be inaccurate. In light of this inaccuracy the prototype was required in order to determine the device’s feasibility. This is due to the fact that calculations are subject to too many unknowns to be confident of the predictions yet research has shown that devices of this nature are effective.
The reason for developing a small scale prototype can also be attributed to construction and time constraints. The design for the full scale apparatus required expensive and difficult to source items. Calculations and advice from professionals included obtaining a honed hydraulic cylinder tube from Perth at a price of $2100, as seen in Appendix B. Also it was advised that a 514 litre pressure vessel would be needed, at a price of $2800 as quoted by CAPS Australia. These parts are very expensive and as such it was risky for the University of South Australia to invest in them without knowing if the apparatus would actually work.
Time constraints were the final factor which contributed to the decision to construct a small scale prototype. As the full scale design required parts which were complex, difficult to work with, and required long ordering times it was decided that a full scale design may take too long to construct, as this project has a finite timeline. Constructing a small scale prototype
required less procurement time, less construction time, less safety design and less set up time. It was therefore more applicable to this project.
The purpose of the prototype is to confirm that the theory is correct, and that this design can be used for a pneumatic cannon of any scale. In particular a pneumatic cannon which is capable of projecting timber planks and steel balls, as defined in the Australian Standard AS 1170.2. The small scale prototype could also answer questions that cannot be revealed mathematically, such as pressure to velocity ratio, and the efficiency of backing wads.
This prototype is an important step towards the full scale design, construction, and commissioning of a wind borne debris testing device which will be created in the future by the University of South Australia.
4.1.3 Timber Projectile
Given the time constraints the main focus of the project has been the timber projectile. The reasons for this are listed below.
 Most of the safety concerns are addressed by the timber projectile
 It is far too dangerous to test with a ball bearing that has a high probability of ricochet without proper safety measures in place. These safety measures will be more refined if timber projectile testing is carried out prior to ball bearing testing.
 The propulsion of the timber projectile poses more of an issue given the large mass compared to the ball bearing.
 It will be easier to adapt the timber projectile device to also fire a ball bearing projectile that to carry out the project in the reverse order.
4.1.4 Ball Bearing Projectile
Due to time constraints this has been omitted for the majority of the project, see above for reasons
4.2 Small Scale Prototype
4.2.1 Objectives
The objectives of the prototype pneumatic cannon are to test the variables in the project and determine any of the necessary information.
The prototype will test the following;
 Relationship between air pressure and velocity
 Most efficient wad size for transferring energy to the projectile
 Requirements of the safety system for the testing apparatus
Testing these variables will aid in developing a full sized pneumatic cannon.
4.2.2 Small Scale Design
Figure 4.1 – Small Scale Prototype
4.3 Components
A summary of all parts required for constructing the prototype is provided in Appendix C.
4.3.1 Barrel
100mm Pressure Pipe (3m Class 9 PVC) was chosen for use as the Barrel of the cannon. The pipe is constructed to withstand a sufficient pressure to provide valuable results. The diameter of the pipe allows for easy use of oversize tennis balls as test projectiles.
4.3.2 Pressure Vessel
150mm Pressure Pipe ((2m) Class 12 PVC) was chosen for the Pressure Vessel component of the cannon. The pressure vessel needs to be able to contain a large volume of air that is to be rapidly released into the cannon. The pipe is rated to 1200KPa, which is higher than the testing device will operate at.
4.3.3 Valve
A Diaphragm Valve was chosen for the use in the cannon. The valve can open in approximately 0.25 of a second. Such a valve will allow a sufficient amount of air to flow through the cannon in a short amount of time. The valve that has been used in this prototype will be the same as the valve used in the full scale model. After much discussion it was decided that best valve for the task would be the diaphragm valve given its much faster opening time. A representative from Norgren advised that although an actuated ball valve would provide a much higher flow rate, the flow provided by the chosen valve would be sufficient to carry out the task. Full specifications for this valve are provided in Appendix D.
4.3.4 Projectiles
Large Tennis Balls were chosen to be projectiles for testing. The soft projectile tennis balls are the ideal size to provide only small clearance to the internal diameter of the Barrel. Using a soft projectile will ensure that when a projectile is fired the safety risk will be small.
4.3.5 Nylon Wad
A nylon wad was tested to provide information for the large scale model where the projectile is not a suitable size for the barrel. The Wad will seal the area behind the projectile in order to provide a transfer of energy to the projectile and make the system more efficient.
Figure 4.2 – Nylon Wads
4.3.6 Instrumentation
Significant instrumentation is required to enable the accurate measurement of projectile speed to enable effective testing. Measuring Velocity
After considering the different options for measuring the speed of the projectile it was decided that the best option was to use light gates. This required an apparatus where two light beams were positioned in the flight path of the projectile 100mm apart. When the projectile is fired it travels along the flight path and breaks the two light beams, with a digital counter measuring the time between each light beam being broken. This time difference is then interpreted by a computer program which calculates the speed of the projectile depending on the distance between light gates (100mm) and the time taken for the projectile to travel this distance. This system was considered to be very accurate (within 0.1% inaccuracy) and was able to give instant results, which is important when conducting multiple tests as it is time saving. It also minimises human error as the results are all digitally recorded and calculated, and the apparatus can be moved along the flight path so speeds can be recorded at the end of the barrel, at impact or anywhere in between. Also this option is significantly cheaper than purchasing the previously mentioned radar gun.
The digital counter, shown below in Figure 4.3, records a count every microsecond and as such, is extremely accurate.
Figure 4.3 – Digital Counter, Records 1 Million “Pulses” Per Second.
As an example, a projectile travelling at 40 m/s would travel 100mm in 2.5 milliseconds. In that time the digital counter will count 2,500 pulses with an accuracy of ± 1 or 2 pulses. Therefore the accuracy of this system is more than adequate for this application. Light Gates
There are two separate light gates which are positioned 100mm apart within the apparatus. These light gates consist of an LED laser which creates a beam that is directed into a light sensitive detector. This detector senses when there is light present, and as such also detects when light is not present. As the projectile passes the light beam will be interrupted and a lack of light triggers the counter to start recording pulses, which therefore records time as the counter pulses every microsecond, i.e. one million pulses per second. The projectile passing through the beam of the second light gate causes the sensor to also detect a lack of light which in turn triggers the counter to stop recording pulses, therefore stopping time recording.
Figure 4.4 – The Two Laser Light Gates Housed Within the Steel Structure, Spaced 100mm Apart.
Figure 4.5 – Circuitry that Interprets “Pulses” Into Digital Data for LabVIEW.
29 Apparatus Structure:
The light gates are housed within a steel frame which was created specifically for this apparatus. The frame enables the system to be easily moved to different heights and can be positioned at any distance along the flight path of the projectile. Also, as the frame is made of square hollow sections there is a minimal chance of damage being caused by the projectiles if they happened to collide with the apparatus. The LED lasers and light sensitive receivers are housed within the square hollow section, as shown above in Figure 4.4.
The apparatus has the following dimensions:
Table 4.1 – Frame Dimensions
Cross Section
Length (mm)
Height (mm)
Width (mm)
Top tube
Leg tubes x 2
Housing for LED’s and Sensors x2
Figure 4.6 – Structure of Instrumentation
Top Tube
Leg Tubes
LED and Sensor Housing
30 Computer Software:
The software used to analyse the digital data was LabVIEW, which is a software program specifically designed for engineering laboratory applications. This program converted the pulses over a time into a speed instantaneously and displayed this result on the computer screen. This software was programmed by the technical staff at UniSA, an example of an output of the program is shown below in figure 4.7:
Figure 4.7 – Output from LabVIEW
As can be seen the software is easy to read and interpret. The output data was accurate to one tenth of a metre per second, which was considered to be acceptable due to the degree of possible variables and error concerned with initial testing. The software is simple to reset and can be used repeatedly. Performance in Testing:
Testing was conducted with the speed measuring apparatus as close to the end of the barrel as possible, as shown below in figure 4.8:
Figure 4.8 – Positioning of the Apparatus
This was kept consistent throughout the testing procedure to eliminate the variable of moving the speed measuring apparatus, and it was positioned as accurately as possible to align the LED lasers with the centre of the barrel so that the centre of the tennis ball will be intersecting the lasers, as this is the point of largest diameter. This is important to keep consistent in order to reduce error due to variables. It was found that when the speed measuring apparatus was used with the projectile and the wad, large variations in speed were recorded. These speeds varied up to 50%, which may have been caused by either the speed measuring device being confused by two objects intersecting the lasers, being the tennis ball and the wad, or caused by the wad being inconsistent with the pressure it captured and therefore the speed to which it accelerated the tennis ball. These two options are discussed in detail in Chapters 6 and 7.
Measurements made with the single tennis ball projectile, with no backing wad, were much more accurate and consistent. These results varied by small amounts of only ± 1 or less metres per second. This proves that the apparatus is applicable for this application and others in the future, such as timber plank and ball bearing projectiles.
Overall it has been established that this speed measurement apparatus is ideal for this application. It is highly accurate, can be used with any form of projectile, and can be moved and altered to accommodate for different testing criteria. It also gives clear and instant results.
32 Reading Pressure
Pressure gauges and regulators have been used to ensure the correct pressure is being distributed to the pressure vessel and projectile.
4.4 Safety Strategy
4.4.1 Mounting the Testing Device
To prevent recoiling of the system the cannon has been mounted to two large steel I-Beams. The 300UC I – beams had holes drilled into them to bolt on saddles to hold down the cannon. The saddles are rubber lined to prevent slippage of the barrel and pressure vessel.
The large beams are very heavy and will act as a counter weight and prevent any movement from the system. Figure 4.9 shows the drilling of the holes and the cannon bolted down in location.
Figure 4.9 – Hole Drilling and Mounted Device
4.4.2 Containing the Projectile
For the prototype testing method, the development of a safe testing area was a high priority. There was little knowledge of how the tennis ball projectile would behave at high velocity. In liaising with the project supervisor and laboratory staff at the University it was predicted that the projectile may shatter or rebound upon impact with a solid object.
Therefore containment of the projectile on all sides was needed. The space between two buildings at the University was used to enclose the sides of the testing apparatus and a temporary wall was built at which to aim the projectiles.
To contain the projectile at the end of the firing range a wall was constructed to absorb the impact, and stop the projectile (Figure 4.10). This wall was constructed using a wooden frame, with a 15mm thick wooden board screwed to it. Glued to this was a layer of 150mm thick, dense foam. The purpose of the foam was to absorb the force of the projectile hitting it to reduce the probability of it rebounding, as well as stopping the projectile from being damaged on impact. The wall was secured upright using clamps and a triangular brace for stability.
Figure 4.10 – Construction and Finish of Wall Barrier
As seen in Figure 4.11, a clear Perspex screen was positioned in front of the pressure vessel, which provides protection from the danger of the pressure vessel exploding. This gives operators a safe view of the testing process and ensures that safety requirements are met, as described in section 4.3.
Testing in progress signs (Figure 4.11) were also placed around the laboratory to notify other people present in the laboratories at the time of testing.
Figure 4.11 – Clear Protective Screen and Testing Signs
Wooden Frame
Triangular Bracing
4.4.3 Noise and Eye Protection
During testing earmuffs and safety glasses were worn. These safety objects were used to dampen the effects of the high noise levels occurred on impact or at the point of fire and to provide eye protection.
Clear Perspex screens also added protection to eyes and loud noise.
4.4.4 Risk Management Plan
A Risk Management Register has been completed for the prototype and included in Appendix E.
The risk register covers the dangers that may be caused by flying projectiles, compressed air and the surrounding testing area.
There have been nine risks associated with the project and ranked on a risk priority level of low, medium and high.
There have been three risks associated with flying projectiles. The highest risks found were to be when the projectile either pierces a specimen or shatters on impact. Adequate controls have been provided for these risks.
There were three risks associated with using pressurized air. The highest ranked risk found was the pressure vessel exploding. Adequate safety controls have been put in place for such an event but it still remains quite hazardous, as the nature and size of an explosion would be unknown.
There were three risks associated with the testing setup; all were considered a low risk.
The safety system that has been put in place to ensure safe operating of the testing apparatus is adequate in controlling the above risks. The use of clear screens, brick and wooden walls and equipment supplied by the University will eliminate any risks for the tennis ball projectiles used for the prototype. Additional measures that will be required for the final projectiles of timber and ball bearings are discussed in Chapter 6.
4.5 Conclusion
The safety strategy detailed above was adequate to ensure that the prototype testing was safe and successful. Changes and additions will be required however, to ensure the safety of testing when different projectiles are tested. Recommendations of these changes and additions are discussed in the recommendations found below.
5.0 Testing of Prototype
5.1 Testing Procedure
In order to determine the most efficient design of a pneumatic cannon, testing of the prototype was performed to determine any variables in the system. In initial discussions it was suggested that barrel length was a variable that may be tested. However, it has since been decided that, given the length of timber projectile required, the barrel will be made no shorter than 3m for safety reasons relating to accuracy. Hence the variables that were tested using the prototype were:
 Pressure & velocity relationship
 Wad efficiency
 Velocity decay
 Instrumentation calibration
The testing procedure that was followed during testing has been outlined in Appendix E. Of the four tests listed, only two were conducted due to time constraints, the wad efficiency test and the pressure versus speed of projectile test. These two tests were chosen as it is important to know the efficiency and ideal size of backing wads to be used for future applications with projectiles such as timber planks, and it is also important to understand the relationship between air pressure and speed of the projectiles.
It was decided to omit the velocity decay test as it was deemed to be irrelevant to know the velocity decay of the tennis ball in the prototype. A timber plank in the full scale model will behave differently due to a much larger weight and having different aerodynamic properties, and as such this test will be undertaken when timber plank testing commences in a future project. Also the calibration of the instrumentation test was omitted. The University of South Australia did not have access to any devices which could accurately calibrate the instrumentation developed in this project. In order to accurately calibrate the instrumentation, further apparatus would need to be constructed. This was not considered viable within the scope and constraints of the project, and as such it was decided not to calibrate the instrumentation at this time. This testing will be undertaken before commercial testing commences using the final device.
These two tests of wad efficiency and pressure versus resultant speed must be conducted so that the speeds of projectiles can be accurately predicted in the future. This will eliminate the need for trial and error to achieve the desired projectile speeds as well as definitively proving the feasibility of this testing method.
5.2 Preliminary Test
The purpose of a preliminary test was to determine that the testing apparatus actually works. This initial testing was conducted using a large tennis ball, with the pressure vessel initially charged with a low pressure of 20 psi. All pressure readings were recorded in psi, as the gauge used on the pressure vessel measured in psi. Once it was established that this process was safe, the pressure vessel was charged to higher increments of pressure until a maximum of 60 psi was reached.
This testing occurred prior to the LED laser light gates being constructed, and as such there was no accurate way of recording the speed at which the projectiles where travelling. For these initial tests rough calculations were performed to get an approximate measurement of the speed of the projectile. This was done by measuring the distance of the projectile’s flight path and the height that the projectile dropped due to gravity over this distance. The following calculations show an approximate speed calculation for the tennis ball projectile at a maximum pressure of 60 psi.
The drop in height of the projectile over the flight path due to gravity enables the time of flight to be found:
?= 12??2
Where, s = Height dropped over flight path = 0.21 m
g = Acceleration due to gravity = 9.81 m/s2
t = Time of flight = unknown (s)
Therefore; ?= √2 × ??
?= √2 ×0.219.8=0.207 ???????
As shown an approximate calculation of the time of flight for the projectile is 0.207 seconds. This is now used to calculate the speed of the projectile:
Where, s = Length of flight path = 8 m
t = Time of flight = 0.207 s
v = speed of projectile = unknown (m/s)
Therefore; ?= ??
?= 80.207=38.6 ?/?
As shown an approximate calculation of the speed of the projectile was found to be 38.6 m/s. This is a very good result as it demonstrates that the maximum speeds can be achieved using this apparatus. However accurate speed measurements were required in order to prove this.
5.3 Results
5.3.1 Wad Efficiency Test
The first test conducted was a wad efficiency test. The Projectile was fired at a pressure of 50 psi with three different wad sizes being tested. The wad sizes vary by 0.5 mm in diameter. This is a very small difference between the diameters of each wad, however this difference is significant and made noticeable changes in the velocity of the projectile due to the way in which they seal the air. The three different diameters for the wad sizes are detailed below:
Table 5.1 – Wad Sizes
Wad Size
Diameter (mm)
Barrel Size
Internal diameter
The three wad sizes have slightly different diameters and as such they will fit differently inside the barrel. The purpose of this is to establish whether a tight fitting wad or a wad with some room to spare will achieve the best transfer of energy between pressure and acceleration, therefore achieving the highest projectile velocity.
Each wad was used in a projectile launch 5 times at a pressure of 50 psi, as shown in the results table below:
Table 5.2 – Wad Efficiency Test Results Velocity (m/s) Test Number 1 2 3 4 5 Average Small Wad
25.4 Medium Wad
28.2 Large Wad
The following graphs show this data for each wad, along with the calculated averages:
Figure 5.1 – Wad Efficiency Results – Small Wad.
Figure 5.2 – Wad Efficiency Results – Medium Wad
Speed (m/s)
Test Number
Small Wad
Speed (m/s)
Test Number
Medium Wad
Figure 5.3 – Wad Efficiency Results – Large Wad
As can be seen in these graphs the results of the wad efficiency test varied significantly. There were variations of up to 50% in the speeds recorded for each different wad size. The possible reasons for this variation are discussed in Section 5.4. However the average speeds for the different wad sizes did give some conclusive results. They approximately showed that the largest wad size resulted in the highest velocity, the smallest was resulted in the slowest velocity and the medium wad size was in between. From this we can see that a tighter fit between wad and barrel results in a higher velocity projectile due to more air pressure being captured.
5.3.2 Second Test: Pressure vs Velocity
The second test conducted was a pressure versus velocity test to establish the relationship between increasing pressure and increasing velocity. The only variable in this test was the increase in pressure, and as such this is the only thing that caused a change in velocity. All other variables remained constant such as the projectile type, and the speed measuring instrumentation and position. There was no backing wad used in this test, due to the variability’s that were encountered in the wad efficiency test. The speed measurement instrumentation was positioned to measure at the end of the barrel to minimise variations in flight path velocities, as can be seen in figure 4.8. Pressure was increased by increments of 5 psi from 20 psi to 60 psi to gather results, with 3 shots being fired at each pressure increment to obtain an average for that pressure, therefore increasing accuracy. The results of this test are shown below in table 5.3:
Speed (m/s)
Test Number
Large Wad
Table 5.3 – Pressure vs. Velocity Results
Figure 5.4 – Average Pressure vs. Velocity Results – Raw Data
As can be seen in the graph above, the relationship between pressure and resultant velocity appears to be close to linear. With an increase in pressure there is a respective increase in speed of the projectile. This relationship can be more accurately shown in the graph below:
Pressure (psi)
Velocity (m/s)
Average Pressure vs Velocity
Raw dataVelocity (m/s) Pressure (psi) Test 1 Test 2 Test 3 Average 20
17.0 25
21.6 30
26.5 35
30.5 40
32.7 45
35.7 50
38.4 55
39.9 60
Figure 5.5 – Average Pressure vs. Velocity – Trendline
This graph shows the relationship between pressure and resultant velocity with a calculated trendline. It shows that the relationship can be calculated, and therefore an approximate speed due to a pressure can be easily determined prior to testing and recording the speed. This is a very significant result, as it proves that this device is suitable for the required application and can be appropriately calibrated for use.
The relationship between pressure and resultant velocity is shown to be:
Therefore: ?= ?−5.370.72 Where: velocity (m/s) = V
Pressure (psi) = p
5.4 Discussion
Two tests were conducted, a wad efficiency test and a pressure versus speed of projectile test. The reason for these two tests was to know the efficiency and ideal size of backing wads to be used for future applications with projectiles such as timber planks, and also to understand the relationship between air pressure and speed of the projectiles. Establishing these two unknowns will allow the speed of a projectile to be calculated and known prior to
y = 0.723x + 5.3726
Pressure (psi)
Velocity (m/s)
Average Pressure vs Velocity Trendline
firing the cannon, therefore eliminating the need for trial and error to achieve required projectile speeds.
5.4.1 First Test: Wad Efficiency
The results from the wad efficiency test were inconsistent and therefore inconclusive. Five projectile shots at 50 psi were fired with each size of wad, and from this an average for the performance of the wads were calculated. As can be seen in the results graph, the individual shots varied significantly, up to 50%. This could be due to a number of factors. Firstly it is difficult to know whether the wads were sitting perpendicular inside the barrel prior to firing. If the wad was leaning slightly in any direction, then it would not be capturing the full potential of the moving air behind it and therefore working inefficiently. This would affect the accuracy of the results as there would be a difference in the amount of air captured with each shot fired. Also if the wad was initially leaning in any direction it would increase the friction between the wad and the internal surface of the barrel, therefore requiring more force and hence pressure to reach speed. It is difficult to determine if these factors contributed to the inaccuracy of the wad efficiency tests, as the positioning of the wad inside of the barrel cannot be seen.
Figure 5.6 – Left Image: Perpendicular wad, Right Image: leaning wad
Figure 5.6 shows the difference between the wads positioned perpendicular to the barrel, and leaning in the barrel. The perpendicular was has minimal friction and full potential of air pressure captured, whereas the leaning wad as points of high friction and inefficient use of air pressure.
To overcome this problem and eliminate this as a possibility of error, the wads could be made significantly longer, preventing them from being able to lean in any direction. This is discussed further in Chapter 6: Recommendations for future / full scale model.
Another possibility for the varying results of this test is in the speed measuring apparatus. The instrumentation was designed for a single object to be passing and interrupting the laser paths, and this single object needed to be of a constant length for the results to be consistent. The addition of a backing wad with the testing projectile may have disturbed the instrumentation and caused the readings to vary significantly.
Although the individual test shots varied significantly, the average projectile speed for each wad size did show some plausible results. It can be seen that with an increasing wad size there is an increase in the resultant velocity of the projectile, on average. This is due to the fact that a larger wad has a larger surface area for air pressure to impose a force on, and therefore accelerates to a higher velocity. There is obviously a limiting factor for this, as
Test Projectile (Tennis Ball)
Test Projectile (Tennis Ball)
there would be a point where the wad is too large and hence friction is too great for higher velocities to be achieved. However this limiting factor was not found, as only three different wad sizes were tested.
5.4.2 Second Test: Pressure vs. Velocity
Results from the pressure versus velocity test were much more conclusive. There is a linear relationship between air pressure and the resultant velocity of the projectile. This result is promising as it shows that we can calculate the pressure required for the projectile to reach a certain speed using the equation shown below:
?= ?−5.370.72 Where: velocity (m/s) = V
Pressure (psi) = p
Therefore this test has been successful, and this information can now be used for the design and construction of a full scale model. However there are some factors which still need to be considered.
Firstly the testing was conducted using a large tennis ball as a test projectile. The relationship equation therefore cannot be applied to the projectiles which will be used with the full scale model. The 4kg timber and ball bearing will need to be tested for their own pressure versus resultant velocity equation to be established; however this initial testing is proof that it can be done, as the relationship for any projectile should be linear.
Also this second test was conducted with no backing wad, due to the issues that occurred in the first test. The second test was conducted with no wad to improve the accuracy and consistency of results after seeing the variability of results when using the wad. Obviously when firing the 4 kg timber and ball bearing as projectiles it will be crucial to use a backing wad to utilize all of the air pressure. Therefore the problems encountered with the backing wad will need to be solved before relationships for these projectiles can be established.
Finally the testing was only conducted within a small range of pressure, from 20 psi to 60 psi. This was due to the limitations of the testing equipment, as the gauge used to measure pressure in the pressure vessel only went to a maximum of 60 psi. At higher pressures the projectile may behave differently, and therefore there may be a different relationship between pressure and resultant velocity. Further testing at higher pressures would be helpful in verifying results obtained in this test.
6.0 Recommendations for Full Scale Model
After successfully constructing and operating a prototype design, observations relating to the required changes for a full scale design have been made. Based on these observations, recommendations have been made below to aid in the future construction of a full scale design.
6.1 Components
Firstly, in order to create a full scale design using the same format as the prototype model, a number of components will need to be changed or altered in order to accommodate the different projectiles. The barrel of the prototype does not have an internal diameter large enough to accommodate the 100mm x 50mm timber projectile that is specified by the Australian Standard. After examining a section of the specified dimensions it is clear that the minimum possible barrel diameter is 114.5mm as anything smaller than this will not fit. It is suggested that PVC pressure pipe may be suitable for the barrel, as it was for the prototype, and simply finding a section size with an internal diameter as close as possible to this value will be acceptable. Another option for the barrel is a material as described in the quote found in Appendix F. The recommendation of the project team is however to use the PVC pressure pipe as reducing bushes and other fittings are readily available which will ensure ease of construction.
With this increase in barrel diameter the volume will in turn, increase. As mentioned earlier in the report, the volume of air must be approximately four to five times that of the volume of the barrel in order to apply sufficient force to the projectile for the entire length. It is advised that the pressure vessel used on the prototype may not be large enough for this increased barrel size, however it may be valuable to test with this first as the fittings will suit the valve already.
As mentioned, the valve used in the prototype will be sufficient for use in the full scale model as well and will simply require different fittings in order to suit the new barrel size and potential new pressure vessel.
6.2 Safety Measures
In relation to safety measures for the full scale device there are some upgrades required to ensure that the system is safe for use.
Firstly, the temporary backing wall was sufficient for the testing using tennis balls as projectiles but it is suspected that it will need to be upgraded in order to withstand impact by the timber projectile. In addition to this, the buildings used to contain the test projectile on the sides of the testing area will require protection when testing with heavy and solid projectiles, as damage may be caused. It is suggested that at the end of the range, a system similar to the crumbed rubber pits discussed in the report are put in place. Along the sides of the testing space it is suggested that sturdy smooth screens such as thick Perspex or bullet proof glass is used to protect operators while still allowing the testing to be viewed. Rebounding may also be an issue with the specified projectiles and as such similar screens will be required at the front of the testing location to protect personnel and equipment.
Another issue that was raised is that the ball bearing may rebound with a large amount of energy and as such will need to be very well contained. It is suggested that a roof on the testing location be considered in order to completely enclose the dangers of testing.
Whilst protecting personnel from the dangers of the pressure vessel is a high priority, the protective method used for this during the prototype testing caused a number of issues. It was difficult to move around in order to load the cannon and check that the system was charged to the required pressure. Also, it was difficult to view the testing taking place, which is of high importance to ensure the device is operating effectively. It is suggested that a more permanent safety system be put in place which is less restrictive to movement around the apparatus. This may be achieved by constructing an enclosure that completely covers the pressure vessel and is strong enough to protect personnel from a rupture or explosive failure.
6.3 Wad Design
As discussed in the report, the test of wad efficiency displayed results with vast differences. This was attributed to a number of things including possible measurement errors and the in ability to ensure the wad is positioned correctly.
In order to rectify the issue of the wad positioning, it is suggested that the wad length be increase significantly in order to ensure that is unable to twist or tip within the confines of the barrel.
6.4 Instrumentation
With regard to the measurement of velocity of the ball bearing projectile, it is unclear as to how effective the current system will be for this. Given that the projectile is very small it is speculated that it may not pass through both sensors and as such the measurement system will be ineffective. Testing and development will be required to rectify this.
As discussed, the timber projectile was given the majority of the consideration throughout this project. In developing a full scale design it will also be necessary to develop a system to carry out testing with the ball bearing. It is suggested that a very similar technique can be applied by using a smaller diameter valve and barrel. It is suggested that the safety system be developed to allow for both types of testing to be carried out in one location.
In order to prevent measurement errors it is suggested that the program connected to the light gate system be modified such that it measures only one result before shutting off. This will avoid “confusing” the system with multiple projectiles (wad and projectile) passing through.
Finally, the Australian Standard specifies an impact velocity and as such it is important for the projectile to be at the appropriate velocity when it impacts the test specimen. This can be achieved by measuring the velocity close to the impact point however this poses the risk of damaging the instrumentation. It is recommended that the velocity decay test detailed in the testing procedure (Appendix E) be carried out and the relationship of distance travelled and velocity decay is found. This test will ensure accuracy of testing whilst keeping the equipment safe.
These recommendations are based on the observations made during the testing of the prototype pneumatic cannon and will assist in ensuring project success when a full scale model is developed.
7.0 Conclusion
After successfully constructing a prototype testing mechanism to perform testing similar to that required by Australian Standard AS 1170.2; Clause 2.5.7 (2011), not all of the originally proposed project outcomes were met. Due to time constraints and the risks involved in constructing a full scale model, it was decided to develop a smaller scale prototype. This enabled testing to be conducted to evaluate the possibility of constructing a full scale model.
Despite the inability to construct the full scale model to test for windborne debris such as a 100 x 50mm piece of hardwood as stated in AS 1170.2, certain conclusions can be made as to how the originally proposed design will perform if it is to be constructed in the future. Most importantly was evidence that the prototype pneumatic canon operated effectively. The assumption can therefore be made that a larger scale pneumatic canon accommodating for heavier and larger projectiles can be constructed and perform in the same way as the prototype.
Also critical in the findings undertaken in the testing process of the prototype pneumatic canon was the relationship found between the effects on the projectile velocity when increasing the pressure to be released on the projectile. As discussed in the results section of the report the pressure and projectile velocity relationship was linear. Therefore when a future full scale model is constructed, it is known that a linear relationship between the pressure and resultant velocity can be found, and therefore pressures required for a given speed can be accurately predicted.
The use of a pressure sealing wad in testing was thought to improve the transfer of energy between pressure and velocity, however it caused great inaccuracy in results. The wad caused great variations in the velocity readings, which could be due to several factors. Recommendations for the development and use of an effective wad system are provided.
Outcomes of this project have certainly been positive and while the sole project outcome of developing a testing apparatus for impacts due to windborne debris in cyclonic regions in accordance with AS1170.2 was not achieved, many necessary preliminary steps have been made. The major conclusion to be taken from this project is that the prototype device that was constructed can be used as a guide in up scaling to the originally intended full scale mechanism.
A concept design of the originally intended device which will enable the testing in accordance with the Australian Standard to be undertaken was achieved as part of the project. This, along with recommendations of how improvements can be made if the project is to be furthered by the University of South Australia, provides a framework to construct an efficient windborne debris impact testing device.
This exciting and largely hands on project has been very beneficial and enjoyable for the project group, and the University of South Australia. All members are satisfied with the project outcomes and look forward to the potential construction of an up scaled testing device by future project teams.
ASTM International, 1996, Standard Specification for Performance of Exterior Windows, Curtain Walls, Doors, and Impact Protective Systems Impacted by Windborne Debris in Hurricanes, West Conshohocken, USA
Australian Bureau of Meteorology, 1974, Severe Tropical Cyclone Tracy, Bureau of Meteorology, viewed 20th August 2012.

Boughton, G.N, Henderson, D, Glinger, J, Holmes, J, Walker, G, Leitch, C, Somerville, L, Frye, U, Jayasinge, N, Kim, P 2011, Tropical Cyclone Yasi Structural Damage to Buildings, Cyclone Testing Station, James Cook University.
Colorado State University (CSU), 2008, Wind Engineering and Fluids Laboratory, Colorado State University, viewed 20 August 2012,

Crandell, J., Marchman, L., McCoskery, D., Rogers, B., Wagner, M., 2002, Wind Borne Debris – Impact Resistance of Residential Glazing, NAHB Research Centre, Upper Marlboro, Maryland
De Scally, F 2008, ‘Historical Tropical Cyclone Activity and Impacts in the Cook Islands’, Pacific Science, vol 62, no 4, pp. 443-459
Emanuel, K 2008, ‘Tropical Cyclones’, Annual Review of Earth and Planetary Sciences – vol. 31, pp.75-104
Government of South Australia Safe Work SA, 2010, Noise; Why Regulations Are Needed, South Australia, Australia, viewed on 3rd November 2012

Government of South Australia, 2010, Occupational Health, Safety and Welfare Regulations 2010 Part 2: Division 10, South Australia, Australia, viewed on 3rd November 2012

Grayson, M, WeiChang, P, Schiff, S 2012, ‘Three Dimensional Probabilistic Wind Borne Debris Trajectory Model for Building Envelope Impact Risk Assessment’, Journal of Wind Engineering and Industrial Aerodynamics, vol. 102, pp. 22-35.
Holmes, J.D. 2011, Wind Loading of Structures, Spon Press, USA
Holmes, J.D. 2004, ‘Trajectories of spheres in strong winds with application to wind-borne debris’, Journal of Wind Engineering and Industrial Aerodynamics, vol. 92, pp. 9-22.
Kathe, E. L, 2001, ‘Recoil Considerations for Railguns’, Magnetics, IEEE Transactions on, vol.37, no.1, pp.425-430, viewed on 2nd November 2012

Kempler, S 2010, Hurricane Katrina, NASA-Goddard Earth Sciences Data and Information Services Center, viewed on 17th August 2012

Standards Australia, 2011, AS/NZS 1170.2 Wind Actions, Standards Australia, Sydney, Australia
Taylor B., 2006, ‘Recoil Experiments Using a Compressed Air Cannon’, The Physics Teacher, vol. 44, no. 9, pp. 582, viewed on 2nd November 2012

Texas Tech University (TTU) 2003, Debris Impact Testing at Texas Tech University, Wind Science and Engineering Research Centre, Lubbock Texas.
Texas Tech University (TTU) 2012, Debris Impact Test Facility, Wind Science and Engineering Research Centre, Lubbock Texas.
University of South Australia, 2012a, Occupational Health Safety Welfare and Injury Management, University of South Australia, viewed on 20th August 2012,

University of South Australia, 2011, Safe Operating Procedure Development, University of South Australia, viewed on 23th August 2012,
< />
Wills, J.A.B, Lee, B.E., Wyatt, T. A. 2002, ‘A model of wind-borne debris damage’, Journal of Wind Engineering and Industrial Aerodynamics, vol. 90, pp. 555-565.
Appendix A – Legal Advice
From: Woodrow, Ben James – woobj006 Sent: Thursday, 30 August 2012 12:39 PM To: Cc: Julie E Mills (Civ Eng) Subject: University Project
To Whom it may concern,
As part of a bachelor of engineering (Civil and Project Management), a team of colleagues and myself have been given a task as part of our final year honours project and I am writing to you with regard to the legality of this project.
Recently, the Australian Standards relating to wind borne debris caused by cyclones has been amended and as such the testing for this standard must change. The clause in question is AS/NZ 1170 2.5.7. This clause states that in order for a product to comply with the standards it must withstand impact from both a 100 X 50mm timber as well as an 8mm Ball bearing. These projectiles are to be accelerated to speeds determined by the regional wind speed multiplier specific to the area of Australia that the product is intended for. In the worst case regions the projectiles will need to reach speeds of close to 160 km/h.
We have completed a large amount of research and decided that the best way to achieve this acceleration is to use a series of valves and compressed air. The device will be built by professionals at the University of South Australia. It will be fixed to the concrete ‘strong floor’ and all appropriate safety measures will be taken in order to ensure that no harm is caused during operation. The building it is to be housed in is secure and protected by swipe card access.
We have recently met with a member of staff at the Holden Hill Police Station and provided with the Firearms Act. By definition it seems that the device we intend to create is an Air Gun. He suggested that we contact the Firearms branch as this is a special case and we are unsure of what is required with regard to licences, permits etc.
Would you advise that we apply for a permit or is it not necessary given the nature of the device?
Any Information you can provide would be greatly appreciated
Thankyou for your time
Ben Woodrow
Uni SA
Ph: 0428 816 103
Appendix B – Global Metals Barrel Quote
Good afternoon Ben,
I did have you correct email written down,
Sorry for the delay in offering.
Thank you for your enquiry on cylinder tube, please find my offer below.
150.0 mm od x 125.0 mm id 1 @ 5000 mm long $ 2100.00 ea + GST.
Ex stock Perth & subject to prior sale.
Please allow 7-10 working days from day of order for delivery to Adelaide.
Please notice that there is a test certification attached with most of the information relating to this product.
I will advise the KPa rating once information is available.
I hope this is of interest & await a favourable reply.
Please do not hesitate to contact me if you have any questions relating to the offer.
David Borovnik
519-523 Grand Junction Road, Wingfield, SA 5013
PO Box 2474, Regency Park, SA 5942
T: 08 8347 1366
F: 08 8347 2966
M: 0403 422 443
Appendix C – Parts List
Parts needed to construct prototype were:
 Compressor
 Pressure Regulator
 150mm Pressure Pipe (2m) Class 16 PVC: Pressure Vessel
 I-Beams (300UC)
 Pneumatic Diaphragm Valve
 100mm Pressure Pipe (1m)Class 6 PVC: Barrel
 Material for Wad
 Materials to construct saddles for housing
 2 inch fittings
 Foam protection
 Plastic Screens
 Safety signage
 Pressure Gauge for pressure vessel
 Equipment for speed measurement
 Projectiles (Tennis Balls)
 Personal Protection Equipment
 Wooden board
 Pallet
 Treated pine (Wall Cross Bracing)
Appendix D – Valve Specifications
Appendix E – Risk Register
Risk Management Register – Windborne Debris Testing Ref The risk How can it happen? What can happen? Adequacy of existing controls Level of risk ranking Risk priority Further action P Projectiles
Projectiles exit the muzzle of the cannon
The projectile can hit a person
Adequate M
Shatter/ Ricochets
The projectile can shatter/ricochets on impact
Pieces of the projectile can fly in different directions and injure an operator
Adequate H
The projectile can pierce the test specimen
The projectile will can go through the specimen and cause danger to the surrounding operators and laboratories
Adequate H
No A Compressed Air
Air pressure leaving the cannon
Pressure will be expelled from the cannon
Can have a force through the air
Adequate L
Pressure Vessel
Can possibly explode
The explosion can injure operators and the laboratory
Adequate H
Can freeze up
Will Delay Testing
Not Adequate L
No T Testing Apparatus
Walking into beams
Operator can walk into the edge of a beam
Cuts, bruises etc Adequate L
Air pressure leaving the barrel
Hearing can be effected Adequate L
Dust from inside barrel
Dust can leave the barrel and enter the surrounding area
Adequate L
Appendix F – Testing Procedure
Windborne Debris Cannon – Testing Procedure
08 November 2012
The aim of the prototype testing is to determine the feasibility of a full scale model as well as determining the most efficient design. In order to this a number variables will need to be tested. In initial discussions it was suggested that barrel length was a variable that may be tested. It has since been decided that, given the length of timber projectile required, the barrel will be made no shorter than 3m for safety reasons relating to accuracy.
The variables that will be tested are:
 Pressure – velocity relationship
 Wad efficiency
 Velocity decay
 Instrumentation calibration
By testing these variables it will be possible to determine the effectiveness of the design as well as providing some valuable information for the full scale version that is to be built in the future.
In order to ensure that the results are accurate and consistent it is important that a standard method is followed. For each of the testing variables the method is detailed below.
Wad efficiency test
In order to determine which diameter of wad allows the best transfer of energy it is important to test the efficiency of each of the sizes that have been produced.
In order to test this variable the velocity measurement system will be positioned at the end of the barrel in a fixed position. A large tennis ball will be used as a projectile and the pressure vessel will be charged to 50 psi (psi is the chosen unit for ease of readability on the gauge face). Each of the wads (small, medium and large) will be tested one after the other. Each wad will be tested 5 times in order to produce accurate results.
Pressure – velocity relationship
To test the relationship between the increase in air pressure within the pressure vessel and the increase in velocity of the projectile the following test is to be carried out.
The velocity measurement system will be positioned at the end of the barrel in a fixed position. With a starting pressure of 20 psi (psi is the chosen unit for ease of readability on the gauge face) the pressure will be increased in 5 or 10 psi increments (time permitting) with 3 firings per pressure setting. The upper limit for this testing will be 60 psi as this is the limit of the gauge and this will provide more than enough data to deduce a relationship.
This test will be carried out using a large tennis ball as the projectile and the most efficient wad as found in the wad efficiency test detailed above
Velocity decay
This test is required to determine how much velocity is lost over the entire flight of the projectile. It is also required to ensure that initial assumptions are correct in saying that the projectile will begin to decelerate immediately after it has left the barrel of the cannon.
This test will be carried out by firing the large tennis ball using the most efficient wad as found in the wad efficiency test detailed above. The pressure vessel will be charged to 50 psi (psi is the chosen unit for ease of readability on the gauge face). In order to test the decay of velocity the velocity measurement system will be placed in three different positions, one at the end of the barrel, one in the middle of the range and one at the end of the range near the foam safety wall. The projectile will be fired 5 times with the measurement system in each of the three locations.
This will provide adequate information to determine the amount of velocity lost over the length of the range which will aid in determining the ideal location to place the test specimen in the full scale version.
Instrumentation calibration
This test is required to ensure the accuracy of the velocity measuring system.
This test will be carried out by firing the cannon at various pressures and comparing the reading from the velocity measurement system against those provided by the speed measuring radar gun that will be aimed at the projectile near to the location of the measurement system.