Explaining SRS – Part 2
SRS in Defence Applications and Replacement for MIL-STD-901D
Equipment designed for defence applications are usually required to pass tests to ensure that they do not fail when subjected to mechanical shocks that could be encountered in transport or field environments.
The most authoritative and widely used standard for Shock testing of defence equipment is the MIL-STD-810 (the latest version of which is MIL-STD-810G with Change 1). Other standards often specified for defence contractors to meet are DEF STAN 00-35, IEC60068-2-27, RTCA DO-160 (airborne equipment) and BV043 (shipboard equipment). Shipboard equipment is also tested to the archaic standard MIL-S-901D, which is still used, although with less and less frequency.
Options for Shock testing in defence applications
It is widely accepted that the most authoritative resource for environmental testing of defence equipment is MIL-STD-810 standard. The tests are specified in great detail with considerable engineering and scientific knowledge applied to the derivation of test methods and meticulous reasoning given for the choice of particular procedures and the associated parameters. The mechanical shock testing requirements in the latest version, MIL-STD-810G with Change 1, is spelt out in Method 516.7 of the standard. Furthermore, majority of other standards mentioned above are either based on or derived (fully or partially) from MIL-STD-810. Therefore, when considering options for defence application testing MIL-STD-810G is a logical place to start.
The standard offers three basic approaches to shock testing:
- Time Waveform Replication
In this method, a representative shock measured in the field is processed and directly replicated on a vibration machine. This method is limited only to environments where shocks are highly repeatable and always of the same type. Furthermore, direct replication of complex real-life shock waveforms, such as pyrotechnic events for instance, is highly demanding in terms of laboratory resources, which is directly translated into extremely high cost of testing.
- SRS Testing
In this method, testing is conducted to ensure the equipment under test is subjected to shocks that meet a specified Shock Response Spectrum (SRS) profile. The required SRS profile is obtained through mathematical transformation of the measured real-life shock waveform. This method is discussed in detail in Part 1 of this article.
Linear, single degree of freedom shock theory shows that two shock events with different time-domain waveform will produce the same damage potential on a test item as long as their respective SRS profiles remain the same. A crucial point to note here is that while each time-domain waveform corresponds to a unique SRS profile, the same rule does not hold in reverse. In other words, it is possible to generate same SRS profile using different time domain waveforms. Therefore, a complex real-life shock waveform can be substituted by a much simplified signal that is calculated to have same SRS profile.
It is this property that of the SRS method enables even complex real-life shock events to be accurately and efficiently simulated under controlled laboratory conditions. Also, SRS profile can be tailored even further to encompass and synthesise a whole range of different shock events into a single test waveform, thus reducing total cost of testing even further.
Method 516.7 of MIL-STD-810G heavily weights towards SRS testing. It stipulates that the required SRS profile should be calculated from data obtained through field measurement or scaled measurements. The standard also gives detailed guidelines about how the measured data should be processed. However, where such data is not available, default SRS test profiles are given for various scenarios.
- Classical Shock Pulses
In situations where field measurement data is unavailable, MIL-STD-810 also allows for testing through classical shock pulses such as terminal peak sawtooth and trapezoidal. This type of testing is very common in all civilian industries (usually carried out to IEC60068-2-27 standard) and is very easily performed in laboratory conditions at a low cost. However, classical shock pulses are derived purely mathematically and as such cannot replicate any real-life shock events – especially not the complex ones likely to be encountered in defence environment (i.e. pyrotechnical). Hence, this represents the least favoured approach and MIL-STD-810 only allows it as a last-resort option.
Therefore, SRS method of shock testing is clearly the most favourable option for defence applications due to its ability to replicate effects of complex real-life shocks under laboratory environment.
As detailed in Part 1 of this article, Austest has considerable expertise in analysing, processing and tailoring of SRS profiles, as well as range of advanced test equipment capable of generating the required signals.
This method has been used – and with ever increasing frequency – since the early 1960s. It is currently used in a range of applications from design of space vehicles to simulating gunfire, underwater explosions, other pyrotechnic shocks, earthquake events and aircraft take-off and landing.
It should be noted that in addition to the MIL-STD-810 standard discussed above, other relevant specifications (such as IEC60068-2-27, BV43, etc.) also heavily gravitate towards the SRS method for the same reasons.
Furthermore, the interchangeability of the time-domain signals makes SRS testing particularly well suited as a replacement for some of the legacy test methods which lack reproducibility, traceability, or are simply unavailable or expensive to execute. One such example is MIL-S-910D, which is discussed in following chapter.
The MIL-S-901D standard is based on an old machine developed by the British and U.S. Navies during the World War II to determine the damage to shipboard items from an underwater explosion. The original machine was dubbed the Lightweight Shock Machine, while the machine for testing heavier equipment was called the Medium Weight Shock Machine. The procedure used has basically remained unchanged since then (!) and is codified in the MIL-STD-910 procedure. Both machines involve a pendulum hammer that hits an anvil to which the equipment under test is mounted. The machines do not actually produce the shocks experienced by ships in such explosions but have been found empirically to produce the same damage potential as typical explosions near typical ships at the time the procedure was developed – which is the main weakness of this method.
As the methods were developed around 75 years ago and both ship designs and anti-ship weapons have changed considerably since then, there is great doubt as to how well does MIL-S-901D simulate the damage causing potential of modern explosions on components mounted on modern ships.
Not surprisingly, naval manufacturers are moving away from specifying shock testing requirements based on MIL-S-901D. Apart from being badly out of date, an additional problem with the standard is that it assumes that all equipment on a ship will be subjected to the same shock regardless of where on the ship it is located. It is obvious that since a ship structure is not infinitely rigid, the shocks transferred to different parts of the ship are not identical and these shocks are not the same as the shock on the hull. Therefore it is far better to tailor the shock test specification according to the part of the ship that the equipment under test will be mounted to using the methods outlined in MIL-S-810G to determine the shock test parameters.
Furthermore, with the tests on lighter items (under 250 pounds), which must be done on the Lightweight Shock Machine, there is an additional problem with MIL-S-901D testing. Since the impact energy hitting the anvil plate is the same regardless of whether the equipment under test weights just 5 pounds or 250 pounds, and given that a 250 pound test item can make up a significant proportion of the total anvil mass, the level of shock acceleration generated by the test varies considerably with test item mass. This is unlike the actual service situation, where typically, the shipboard structures to which the equipment under test are mounted are so heavy in comparison with lighter equipment that the base acceleration transferred to the equipment will be largely independent of equipment mass (for items light enough to be tested on the Lightweight Shock Machine).
However, the biggest problem with tests conducted to MIL-S-901D is the problem of repeatability. This arises because, unlike all the other shock test standards referred to above, MIL-S-901D is purely a machine-specified standard. In other words, MIL-S-901D does not require shock of a specific severity and type to be generated. Rather the test is specified purely by defining the design of the shock test machine and the procedure that must be followed in conducting the test. The damage causing potential generated by such machines varies a great deal since the peak shock transmitted depends much on the level of damping of the machine and this in turn changes considerably with bolt tightness and friction between surfaces. The problem becomes more acute at higher frequencies, meaning that the test becomes less accurate for less massive items like electronic equipment, whose key components often have critical frequencies about 200 Hz. To add to this problem, as there are no shock severity parameters for the test specified and testing houses are not required to measure the shock imparted to the equipment under test, manufacturers having their equipment tested by MIL-S-901D machines will usually have little idea whether the shocks imparted on their products are typical or on the high or low ends.
Thus shock tests conducted on one MIL-S-901D machine can produce different results than a nominally identical test on another machine run by a different testing house. This is because wear of the bearings and other mating parts, changes in bolt tightness, deformation of parts and other changes in the machine with use can have a significant effect on the shock loads. The problem is especially acute with the Lightweight Shock Machine because the anvil is known to work-harden with use, which tends to make the machine generate higher shock loads, of higher frequency content, over time.
Think about this nightmare scenario. You get MIL-S-901D testing performed on your product on a Lightweight Shock Machine. However, in service, the product suffers damage and your end customer is not happy with you. A commercial dispute arises and the reputation of your firm is under threat. You point out that the equipment had previously passed MIL-S-901D testing. To check that you actually had the equipment properly tested, the customer then gets the product MIL-S-901D tested by a different testing house. However, the other machine has slightly different bolt tightness/friction/level of deformation/level of anvil work hardening. It then turns out that in this latest MIL-S-901D test, your equipment actually fails the test. Now you are in big trouble!
If that is not bad enough, consider the following scenario which could make this average nightmare turn into a scene from your worst horror movie: You insist to your customer that the equipment was properly tested originally and that the new testing house may have done the test incorrectly or have a faulty machine. So you go back to the original testing house and have the equipment re-tested. However, since your equipment was originally tested, the anvil in the Lightweight Shock Machine at this original testing house has work hardened further through repeated use. So the imparted shocks are higher … and now your product fails the test!! What will you do now?
Therefore, for the many reasons listed above, we strongly advise manufacturers of shipboard equipment to move away from getting their products certified to the archaic MIL-S-901D and instead seek testing according to the more precise MIL-STD-810G or other properly specified standards. However, should end customers insist on to MIL-S-901D testing, we advise equipment manufacturers to have shock testing conducted to meet an SRS profile calculated from acceleration-time data recorded from past MIL-S-901D tests. Austest can conduct such SRS testing based on one of the various methods outlined in Part 1 of this article. We even have the capability to produce an SRS typical of the Medium Weight Shock Machine.
By replicating the shock response envelope produced by typical MIL-S-901D tests through scientific SRS testing, our testing will ensure that such tests are repeatable and fully specified.
Authors: Praba Naganathan and Goran Tomic, Environmental Test Engineers, Austest Laboratories, Copyright 2016