You need real-world tests to complement finite-element analysis
Most electronics engineers don't give much thought to strain measurements—EEs are usually more concerned with electrical reliability than with structural integrity. But for engineers who develop products such as aerospace components and semiconductor-processing equipment, strain is a crucial measurement. If a design doesn't hold together structurally, it doesn't matter how well the electronics work.
An increase in the number of composites and the rising sophistication of structural and mechanical components keep expanding the need for more accurate strain measurements. While traditional stress-strain equations may work fine on cylinders, beams, and other structures made of homogeneous materials, they fall short for irregular shapes and nonhomogeneous composites.
Moreover, as manufacturers look for ways to shrink the size and weight of their products, parts are being stressed closer to their strength limits. This reduces the margin of error below that which calculations or computer modeling can achieve alone, and it raises the need for a very precise physical measurement.
Demand for strain-measurement resolution to the fraction of a microstrain range (millionths of an inch per inch) is on the rise in design and prototyping of autos, medical devices, semiconductor-processing equipment, and sports equipment. Airframes or truck chassis that were once strain-tested at 50 measurement points and to 10 or 100 microstrain are now monitored at 150 points to the single-digit microstrain level or below.
Consider two examples in which we at HBM have been involved. In the first, a semiconductor crystal furnace uses strain gages on a load cell to measure weight change of the crystal as it grows. It requires strain measurement resolution of 1.06 microstrain. In the second, the rim of an aircraft landing gear needs strain-measurement resolution in that same range to verify a new lightweight design. With varying wall thicknesses to optimize the strength-to-weight ratio coupled with a very irregular geometry, the shape of the rim defied any type of modeling or calculations to quantify stress at critical points. Using high-precision strain measurement, the designers were able to shave several extra pounds out of the part.
Designing the test setup
How can you be confident of getting good results in high-precision strain measurements? First, decide what you want to learn from your test. This may seem self-evident, but often, it isn't.
For instance, a designer of ceramic hip implants recently called us to order a strain gage and amplifier to measure strains in a prototype socket. I asked which direction of strain he was most interested in, and he simply didn't know. We finally settled on a rosette, which measures strain in three directions at once.
Once you know what you want to learn, you need to follow three basic steps to set up your strain-measurement system:
- Select the right strain gage,
- Select the right amplifier, and
- Install everything properly.
If, as in the hip-joint case, you have a complex project, be sure to ask for help early on. You'll save a lot of time, money, and false starts.
Choosing strain gages
For single-microstrain-level measurements, the only real choice is the foil-type resistive strain gage. Although semiconductor strain gages work fine under stable conditions, they are subject to reading errors from many sources, especially changing temperatures and even variations in light and darkness.
Foil-type strain gages work on the principle that a change in length changes the resistance proportionately. In an energized Wheatstone bridge, that change in resistance generates a highly precise, very stable, proportional signal. The resistance material is usually constantan, but it may also be CrNi. Strain gages are also available in a variety of resistances, lengths, and geometries.
These eight guidelines cover the most frequent problems we see in the field:
- For composites and other nonhomogeneous materials, select the longest practicable strain gage. The greater length increases the odds of capturing a true average strain for the material as a whole.
- For high-humidity applications, specify the lowest practicable resistance. The lower resistance offsets errors caused by the swelling of the insulation in response to humidity.
- For strain measurements around irregularities and stress-concentration areas, use the smallest practicable strain gage. You'll get a truer strain reading at the worst-case point.
- For ambient-temperature applications, constantan-based strain gages work fine. For temperatures above 200°C, specify CrNi gages. And be sure to consider the temperature of the surface where the gage is mounted, not ambient air. Even on a 72°F day, metal gets hot in sunlight.
- Use CrNi strain gages for long-term cyclic strain tests. CrNi delivers a longer fatigue life than constantan.
- Use encapsulated gages. Encapsulation protects the gage against handling damage.
- Use gages with built-in leads. This saves a lot of soldering and installation time, and it reduces the risk of damaging the gages during installation.
- If you know the direction of principal strain, use uniaxial gages. If you're not sure, go to a rosette or side-by-side gage. Both measure axial, biaxial, and shear strain at a point.
Look at amplifier specs
As important as it is to choose the right strain gage, the choice of an amplifier is even more critical for achieving high precision. Strain measurements can be no more accurate than the amplifier that processes the signal from the strain gages (although even the best amplifier can't undo mistakes in strain-gage selection and installation).
If you compare amplifier specifications carefully, you'll find big differences—differences that can affect performance substantially. Here are some tips for finding the best amplifier for your application:
- The key specifications are bus frequency and resolution. A high bus frequency enables you to capture transient events better. The resolution determines how precisely a data-acquisition system captures the analog strain data from the strain-gage signals. It is the amplifier's resolution characteristic that most influences the precision of strain measurements.
- For highest measurement accuracy, look for 24-bit resolution and high bus frequency. For example, a 19.2-kHz amplifier can fully describe a 1-ms transient in an airframe that a slower amplifier would miss completely. And most transients in transportation structures happen that quickly.
- For versatility, look for an amplifier with both carrier-frequency and DC operating modes. Whenever possible, try to run your strain tests in the carrier-frequency mode. This mode spontaneously eliminates a lot of measurement error sources, such as electrical and thermal noise and losses over long lines, and it creates a purer signal for data storage and analysis.
- Be sure the amplifier can handle two-, three-, and four-wire hookups (Figure 1). These circuit arrangements with the strain gages can improve accuracy where long cables between gage and amplifier cannot be avoided. (See more about cables under "Installing a system," below.)
- Compare robustness. Available amplifiers vary widely in this regard. If possible, pick up the amplifier and see how solid it feels. Look at the frame construction. Slide cards in and out of their slots. You want an amplifier that will stand up physically to the testing environment for many years.
- To simplify setup, look for an amplifier with built-in displays and controls. You'll be able to check things out as you go right at the amplifier and won't need a PC at that stage.
Installing a system
Though reasonably rugged, strain gages do need to be handled with care. In particular, keep them flat, and don't put anything on top of them. Here are four other things to keep in mind during installation:
- Be aware that long analog cables between the strain gage and the amplifier are susceptible to electrical noise and line losses, and this can introduce errors. Keep cable lengths as short as possible. Consider using three- or four-wire hookups as well as intermediate analog-to-digital converters.
- Surface preparation is essential for obtaining a meaningful strain measurement (Figure 2). Be sure the surface is absolutely clean and smooth (we recommend a 16-microinch finish), so the gage can directly contact the material of interest. Otherwise, you may not be measuring what you think you are. Any tool marks or similar irregularities can deform the gage enough to destroy it or introduce errors. We recommend getting down to bare material, then roughing it slightly with sanding or grit blasting.
- Room-temperature curing adhesives are convenient and are adequate for shorter term testing at normal temperatures. But if you are planning extended tests lasting multiple months or if there's any chance of heating at the point where the gage is installed, use heat-cured adhesives.
- When connecting leads, be careful not to overheat or damage the strain gage. Also, examine joints to ensure they have been properly soldered; poorly soldered joints can introduce resistive errors.
Even with the proper equipment in place, you need to be watchful for any factors that could affect your measurements. Recently, we helped one of our customers prepare an airplane wing outdoors for load and strain testing that was to start the next day.
All strain gages were in place, properly wired to reduce analog cable lengths, and all connected to a precision measurement amplifier. The next morning, the system registered a change in strain, even though no loads had been applied. It turned out that we had measured strain in the wing due to the weight of the dew that had settled on the wing overnight.