Cross Webinar – Testing 101: Vibration Testing

Vibration and Shock Testing

We’re going to go into some guidelines of testing as well as control strategies and fixturing.  For those of you who don’t know me, my name is T.C. Beinke and I’m the national account manager for the testing division here at J.A. King.

What is Vibration?

Vibration is defined as the periodic motion of a rigid or elastic  body around an equilibrium point.  And when we’re talking about vibration testing, specifically, that body that experiences the vibration we refer to as the DUT, or device under test.  And typically with vibration testing, we’re going to be testing under one of the three mutually perpendicular axes, X, Y or Z.

Why do we perform vibration testing?

We do this to simulate stress for real world vibration scenarios. This could be vibration from an automobile going down the road experiencing rough terrain. That’s a good example of a real world vibration scenario and I put some images here showing some other examples as well. We could be talking about vibration from machine operation, vibration from a locomotive or vibration from aerospace components and even wind itself can cause vibration.

Vibration Vocabulary

I also want to go into some basic vocabulary about the shaker system itself. So you’ll see here I’ve pointed out a shaker in the horizontal and vertical orientations. The armature is the moving member of the shaker and this is what’s actually going to transmit vibration into your fixture and your DUT . The armature is going to be bolted to the head expander or the slip table, whether you’re in the horizontal or vertical orientations. The point of the head expander or the slip table is to give you a larger working area to bolt your fixture or DUT to for vibration testing.

Main Profile Types

There are a few different types of profiles both for vibration and shock. For our purposes today, we’re going to stick with the ones that are most common that we run into all the time. For vibration, that’s sinusoidal and random vibration testing. And for shock, we’re going to be talking about classic shock testing.

Sinusoidal Vibration

Let’s start with sinusoidal or sine vibration testing. This is the simplest form of vibration and the easiest way to think of sine vibration is to consider a mass and spring system. This mass as it’s supported by the spring is going to operate along a sine curve. You can see here that I’ve put in a figure of a basic sine curve and pointed out some of the key parameters. On the right side, I’ve pointed out the displacement which is the distance from the x axis and the peak of that sine curve. A lot of times when we are talking about sine vibration testing, we refer to the peak to peak amplitude which is just double the amplitude itself. Also, I want to point out the period which you can see on the left. We also talk about the frequency a lot and the frequency is one over the period or the reciprocal of the period. Also, this is an example of a sine dwell and all the parameters in this case are staying constant. The displacement, the frequency and the acceleration are staying constant as time goes on and nothing’s changing. We are going to talk about a sine sweep in a little while and you’ll see that that is not the case there and we will have some parameters changing.

So why do we use sine vibration?

The main reason we use sine testing is to find the resonant frequency of a system. A natural frequency of a system is the frequency of the system oscillates when it’s not subjected to a continuous or repeated external force. It’s the frequency that the system is going to oscillate at when it doesn’t have an external vibration applied to it. Resonance occurs when the applied frequency equals the natural frequency. It results in the exponential increase in the amplitude of vibration. If you take a look at the figure I put at the bottom, this is a good example of what a resonance frequency would look like graphically. You can see as the frequency increases and you do hit that resonant frequency, around about 175Hz, you get a huge spike in your vibration amplitude so looking at this graph, you would know that the system you are testing has a resonant frequency of 175Hz.

Usually when you are doing sine vibrations, you are going to perform a sine sweep. A sine sweep is used to find resonance in either the DUT itself or the fixture. When you want to find resonance in the DUT, you usually want to find out how much damage you will experience at that resonant frequency. So usually you’d do a sine sweep over a specific frequency range and once you find that resonance and then you’d perform a dwell on it to see what sort of damage your DUT is going to experience.

The other case would be doing a sweep for fixture validation. The reason that’s important is to make sure the resonant frequency of your fixture does not fall within the frequency range of the actual test that you want to use that fixture for. If you look down here at the bottom, this is just an example of time domain data of a sine sweep. You can see as time goes on that frequency is increasing.

This is an example of what a sine sweep would look like on the actual vibration controller. In this case and in a lot of cases with sine sweeps, you’re going to start out with an increasing acceleration and a constant displacement until you reach your target acceleration, in this case 1G. Once you reach that target acceleration, you’re going to continue at 1G all the way up through your frequency range and back down. I also want to point out the axes, you can see that both the X and Y axis are logarithmic. The reason this is important is because it’s going to enlarge small acceleration contributions which will help them show up on the graph so you don’t miss any spikes in acceleration that way. Also, I want to point out that the x axis is frequency and it is proportional to time. If you were to perform a sine sweep, you’d actually be able to see your shaker operating along that curve and it would track along that green profile line, from the left side of the screen, up and to the right side of the screen to the higher frequency range. You’ll see that with random vibration, the x axis is not proportional to time.

Random Vibration

Random vibration is used to simulate a real world vibration scenario. Whether it be transportation or machine operation, any of those real world vibration scenarios that we talked about at the beginning of this presentation, that’s what you would use random vibration for. It’s much more realistic than sinusoidal and that’s because in the real world, the future of vibration is not predictable so you need something much more random and sporadic to actually accurately simulate it. You can see here, down at the bottom, that I’ve put a figure of time domain data for random testing and you can see that it’s much more sporadic, especially when you compare it to a sine graph.

For sine testing, the intensity is dictated by your G level or your actual acceleration. For random vibration, it’s different and your intensity is dictated by the Grms and that’s a measure of the cumulative energy input or a measure of how your shaker is working. To give you a better idea of what is really the Grms, it’s calculated as the square root of the area under the profile curve so you can see that this is what a random vibration profile would look like on your controller and the square root of the area under that green curve is going to be the Grms. It is important to note that while this Grms does indicate your test intensity, two different curves with the same Grms are not necessarily equivalent. They could have completely different shapes, and different frequency ranges as well.

I want to talk about the x and y axis for random as well. As I mentioned earlier, the x axis here is not going to be proportional to time, so rather than tracking along that green profile curve like it did with sine testing, with random vibration, at any point in time, your shaker could be operating anywhere along that curve, so it will actually bounce around that curve as you’re performing testing.

Also, I want to point out the y axis. You can see for sine testing it was just Gs, your acceleration. For random testing, it’s (gn)²/Hz. These units confuse people sometimes but essentially, if you take your acceleration and square it and divide it by the frequency, you’re just normalizing those values. This is important because the acceleration of random vibration testing depends upon the bandwidth that you measure that acceleration on.

Before you perform a test, the profile itself needs to be defined and for random vibration, there’s a few different ways of doing that, both graphically and by using a break point table. You can see here on the left, this is an example of a graphic representation of a random vibration profile, which speaks for itself. It gives you a graphical representation of the actual shape of the curve itself and gives you some key frequencies and accelerations. In this case, they also gave you some slopes as well.

A break point table, on the other hand, instead of a graphical representation, it gives you every single key frequency point and its corresponding g²/Hz. When you look at vibration specifications and procedures, you’ll see a mix of this or one or the other.

Mechanical Shock Testing

I’m going to move on to mechanical shock testing. Shock testing is essentially used to simulate impact, drops or some sort of explosive force. The reason shock testing is so important is that shock excites all resonances. We’ve already talked about how important it is to test at resonant frequencies because you’re going to get that increase in amplitude and when you perform shock testing, each component of what you’re testing is going to quiver at its own natural frequency. You can see here that I’ve included a figure of a half sine pulse and I’ve pointed out some of the key parameters here. You can see there’s the amplitude at the top of the curve which is the acceleration level. I’ve also pointed out the pulse width which is another key parameter. Pulse width is the duration of the test itself.

You’ve already seen an example of a half sine pulse and this is the most common pulse shape. On the right, you can see what that pulse would look like in a vibration controller. There are other pulse shapes other than half sine, such as sawtooth, trapezoidal and rectangular. The sawtooth shape is what you see here at the bottom. As I mentioned, half sine is the most common but these days, specifications are moving towards sawtooth because there is a more realistic residual spectrum for sawtooth pulse shapes.

Climatic Vibration Testing

I also want to talk about including a climatic element in your vibration or your shock test. This can be especially useful for transportation tests. If you think about having your car in the back of a truck, it’s going to be bouncing around, it’s going to experience vibration and shock, but it’s also going to have some environmental conditions that are changing during this as well. By including a climatic element in your test, it’s going to more closely simulate a real world environment. This is really important because temperature and humidity can affect the material properties of your part. So if you’re testing at a higher temperature, the material is going to be more ductile and at a lower temperature, it’s going to be more rigid. This could affect the results of your test.

This is an example of a setup at our Greenville, SC facility. You can see we have our shaker and right behind it is our environmental chamber. You can slide the entire vibration system into the climatic chamber, seal it up and perform vibration testing with the climatic element, whether it be climatic aging or temperature cycling.

Common Testing Specifications

I also want to list a few common specifications that we run into a lot for vibration and shock testing. For transportation, I have listed a couple of different ASTM specifications. Both of these we run into quite a bit and both are quite useful because they give different profiles based on the method of transportation, whether it be air, ship or truck transportation.

Also, I pointed out a couple of automotive specifications for vibration testing. That ISO specification is very useful. It’s geared more towards automotive electronics but it is often used for all types of automotive components, other than electronics.

I also pointed out an IEC specification. It’s a great spec and provides a lot of great guidelines for vibration testing but it’s important to note that it’s a guideline specification rather than providing actual test parameters. The IEC specification, as well as the military standard I highlighted here both will require some tailoring on your end.

Of course, OEMs will have their own specifications for vibration as well. Often, they will reference back to one of these other specifications that I’ve already listed.

Measuring Acceleration

How do we measure and control our acceleration during testing? We do that with accelerometers. The most common type of accelerometer is piezoelectric, which is what we tend to use. You can see I put a figure here showing the basic makeup of how these accelerometers work. It’s essentially a mass and spring system. The spring is a crystal and as that crystal compresses or elongates, it generates an electrical charge which is then converted back to acceleration by your vibration controller. The reason why we like to use these piezoelectric accelerometers is because they have very high resonant frequencies. That means you can use them for a wide range of vibration tests. One of the main downsides to these is that they can’t measure very accurately at low frequencies. Pretty much anything below 5Hz they won’t be very accurate but luckily, most vibration tests are above that level.

If you do want to measure and test below 5 Hz, you can use a capacitive accelerometer. There are some downsides to those accelerometers. For example, they aren’t great at the higher ranges. We tend to stick with the piezoelectric accelerometers but you do have to choose your accelerometer based on the actual test you are doing.

Accelerometers are only as good as where you place them. You want to make sure you are controlling your acceleration as closely as possible to the input vibration location so you want to place it on the fixture itself, not directly on the table and you want to have it near a mounting location and on a rigid member. One exception to this is performing transportation testing because for these tests, you usually don’t have a fixture. You’re just placing the DUT directly on the table and letting it move about freely during testing. In this scenario, you would place your accelerometer directly on the table for control.

Speaking of accelerometers and controlling vibration with them, there are a few different control strategies that can be used. Essentially you can either control with one accelerometer or with several. The most common control strategy is just single control, which uses only one accelerometer, which is what we usually use.

You can also use multiple and there are a few different ways to do that.

• Maximum control is based on the accelerometer that is experiencing the highest level of acceleration
• Minimum control is based on the accelerometer that is experiencing the lowest level of acceleration

The reason you would want to do that is to make sure that no point of your DUT experiences a level above, for example, 5G, if you’re doing a 5G test. You could place multiple accelerometers around your DUT and use a maximum control strategy and that way, you can ensure that no part of your DUT will go over that limit that you have set. The downside of this is that you will have some under-test, some parts of your DUT will experience under-test and if you’re using minimum control, you’ll experience the opposite. You’ll have some parts of your DUT which will experience over-test.

Another option if you’re going to use multiple accelerometers is to use an averaging control strategy. That takes a weighted average of multiple accelerometers which is useful for minimizing the spikes in acceleration but it’s important to keep in mind that an averaging method cannot correct for improper fixturing.

Fixturing

A fixture is really just anything that couples your DUT to the armature, the head expander or the slip table. They can be extremely complicated or as simple as using clamps.

The first question to ask when designing a fixture is what type of material should you use. A lot of people think that the best route to use is steel because it’s rigid and it’s strong. This is not the case for vibration testing for a few reasons. Steel is very heavy and when you are performing your vibration or shock testing, you want to keep your weight as low as possible. Also, steel has very poor damping characteristics. If you think of hitting a steel object with a golf club, you’ll have that feedback through your club and back into your hands. Steel performs very similarly during vibration testing, it’s going to ring. If you do have to use steel, only use it with very low frequencies and for vibration profiles that aren’t very aggressive.

Instead, use lighter materials with better damping characteristics, like aluminum or magnesium. Magnesium is probably the best option. It has the best damping characteristics and it’s very light but it can be pretty expensive.

A good middle of the road material is aluminum. Aluminum is what we use in our lab. It is a good deal cheaper than magnesium. It is a bit heavier than magnesium but compared to steel, it’s much lighter and still has good damping characteristics.

Next, figure out how to join the fixture together. The best option in terms of preventing relative motion is a fully cast fixture. Unfortunately, that’s not always the most realistic option from a manufacturability standpoint, it’s not the easiest way to make your fixture. You also can’t change up the configuration for multiple DUTs and you can’t do that with a fully cast fixture.

A welded fixture will prevent relative motion quite well but again, you’re going to run into the problem of changing up the configuration. It’s going to be hard to use a welded fixture for multiple DUTs. A welded fixture is easier to manufacture than a full cast fixture. You want to keep those welds to be very thick. You want a bead size that is equivalent or thicker than the material thickness than you’re using for your fixture.

The next best choice is to bolt your fixture together. Technically, it’s the poorest choice when it comes to preventing relative motion but it’s your best choice if you want to use a fixture for different DUTs. It’s what we use a lot of times. If you’re going to use a bolted fixture, you need to keep these guidelines in mind:

• You want a very tight bolt spacing. Reduce the distance between bolts as much as possible to maximize the resonant frequency of the fixture itself
• Counterbore all holes to reduce the free length of the bolt itself. This is important because during vibration testing, the free length of bolt will actually stretch a little bit during testing which will allow some relative motion. You want to counterbore all holes to reduce that length as much as possible.
• Since we’re using relatively soft material, you might be concerned with bolts pulling out of the fixture itself. By using stainless steel inserts, you’ll have something more rigid to thread your bolts into.

These are a few examples of tried and tested vibration fixtures that we use at J.A. King. The one on the left is a bookend fixture. Bookend fixtures are very useful because it is a relatively simple design. They are quite rigid and you can use them for a lot of different DUTs.

The one in the middle is a basic cube fixture. Cube fixtures are also very useful because they are rigid and have simple geometry. Cubes are very useful because you can use all the faces of that cube. If you have vertical and horizontal testing that needs to be performed, frequently you have to rotate the entire drum of the shaker. If you are using a cube, you don’t have to do that. You can just bolt your cube in the vertical orientation and then you can move your DUT around to different faces of the cube to reach all three orthogonal axes without having to change up the orientation of the shaker itself.

On the right is another example of a cube fixture but the difference is that some holes have been bored out of the cube itself. The reason for this is to reduce the weight. Those cube fixtures can get pretty heavy, pretty quickly. Unfortunately, once you bore out holes, you will run into some problems with rigidity. A good trick to keep up your sleeve is to squirt damping foam into the voids of the cube which will increase the damping properties of the fixture. It’s important to keep in mind that it’s not going to increase the resonant frequency of the fixture but it will decrease the amplitude of the response when and if you do hit the resonant frequency of your fixture.

I want to point out a few guidelines when you’re designing fixtures. You do want to avoid clamping when possible. I mentioned clamping earlier in the presentation. We do use clamping sometimes and you can use it. You just want to be careful with clamping. Make sure you use it mostly for tests that aren’t very aggressive and at low frequencies. If you do clamp, take note of the exact placement of those clamps. That way, you’ve got a good, repeatable test and if you need to rerun anything, it’s easy to do so in the exact same configuration that you ran it initially.

You also want to make sure you keep your center of gravity as centered over the shaker and as low as possible. This will avoid causing any movement on the actual shaker.

J.A. King’s Vibration Capabilities

This is a quick rundown of our capabilities at J.A. King’s Greenville, SC testing facility. You can see I’ve pointed out the force peaks for both shock and vibration, as well as acceleration for both shock and vibration. Also, I’ve listed our max velocity. For displacement, that 1.5” is per direction so if we are talking about peak to peak displacement, it would be twice that, 3”. Our max frequency is 3KHz and our head expander is 4 x 4’ which is the same size as our slip table. If we need to reduce weight for your test, we can create a fixture that can be bolted directly to the armature rather than using the head expander at all.

Q&A

What is the max payload of your shaker? The max payload of the shaker for shock and vibration is a bit more complex question than you might think. The weight of the DUT+Fixture combination is not the only parameter to consider. It also dependent on the profile that you want to run. I like to say that if you know the weight of your DUT and your profile, you can always send that information to us and we can put it into our controller’s software, and we can get a green light on whether we can run your test. If your fixture and DUT are over what we can perform with that specific profile, we can talk about changing the fixture configuration to reduce weight or we can take out the head expander and bolt directly to the armature.

What temperature range can you run your vibration testing at? We can go up to 170 degrees C and we can go as low as -40 degrees C for both shock and vibration.

Do you guys do your own fixture design? Yes, so if you have your own fixture and you’d like us to use that, we can. You can send it to us and we can validate it by doing a sine sweep to make sure it will work for the test. Or, if you’d like us to completely handle the fixture design ourselves, we can do that, too. You can send us CAD models or drawings or even the part itself, we can design the fixture and perform all the fixture validation that’s necessary before testing.