Temperature Effects on Strain Gage Measurement


The strain measurement is a very straight forward method in experimental stress analysis. The strains obtained from an experiment at room temperature are considered to be accurate, and actually it is accurate with some minor errors if it is performed at unchanged environmental condition. The case arises when the temperature changes, and the specimen is subjected to a different temperature during performing the test. In this case, the strain reading will be affected by several factors that cause error in the reading. When the temperature changes, the specimen will expand, and this expansion causes and increases or decreases in strain reading. The temperature change also affects the gage itself, and the gage can’t read correctly.

This strain change in the specimen and the gage is called thermal output. The change in strain gage by temperature causes the gage factor to be changed. This variation of gage factor is also another factor of error. The error that is caused by different factors can be fixed either by error correction or compensation. There are several methods that can be used for error correction or compensation.  The purpose of this experiment was using different compensation and correction methods to obtain actual strain. The experimental procedure and details for each method will be discussed here.

Experimental Method

A 6061-T651 type aluminum cantilever beam, with specified dimensions in Figure 1, was used in this experiment. Firstly this beam was loaded at room temperature to obtain the strain as baseline strain. Then, four different methods of test were performed in this experiment. Two tests were performed for the correction purpose. Two tests were performed for the compensation purpose.  The other two methods were used to show that there isn't any way to correct or compensate them. These were using two-wire connection and using wrong strain gage. 

Correction Methods

Single beam, 3-wire (simple)

A single beam with a strain gage, as presented in Figure 1, was tested by changing temperature. The temperature was increased from room temperature to 180F˚. The thermal output is calculated from the following equation.

 where T is the increased temperature, and A is coefficient for the equation. The gage factor for this equation is set to 2.0, and the coefficients for an advance (constantan) gage are as following

Since the gage factor is 2.0 and the gage factor for P-3500 was 2.055, there is a correction for thermal output as following.

 Therefore, the correction for the strain can be calculated as following.

The result for the correction is presented in Table 1.

Single beam, 3-wire (more correction)

It is already mentioned earlier when the temperature changes, the gage factor in the gage also will be affected by the temperature change, so there should be an advance correction method to correct the gage factor change or correct the error due to strain gage change. Therefore, there is an advance correction method to correct this error. The correction for the gage factor can be obtained from the following equation.

where FT is the corrected gage factor at increased temperature, FRT is the provided gage factor by
manufacturer which is 2.055,  and delta F(%) is a factor obtained from Figure 7 in TN 504 for the gage
factor change. This factor was obtained 0.7% at 180F˚ for A-alloy strain gage.
Combining equations (4) and (5), the actual or corrected strain can be obtained as following.
where Fis the gage factor of P-3500 which was 2.0. The result is given in Table 1.

Compensation Methods

The compensation method is a method that does not need any kind of equation to correct the error, but there are some techniques to the wiring of the gages to compensate the thermal output. Two techniques were used in this experiment to compensate the thermal output, self-temperature-compensated strain gage and half-bridge with dummy arm.

STC Single beam, 3-wire

A self-temperature-compensated strain gage is a gage that can compensate the thermal output to the possible minimum. This gage uses a 3-wire connection to the P-3500 as quarter-bridge, so bridge can compensate the thermal output at a range of temperature change. A STC, 3-wire strain gage was used on a single cantilever beam, the same as previous parts, to read the strain change by changing the temperature. This reading is also presented in Table 1. This experiment doesn’t need any calculation because of its self-temperature-compensation. Figure 2 shows the diagram of quarter-bridge with a single-self-compensated strain gage.

Half-bridge (dummy gage)

Half-bridge with active and dummy arms and full-bridge with active elements are the methods that can be used as temperature compensation. There are five methods to build a fully temperature compensated Wheatstone bridge. There was only one method used to fully compensate the thermal output effects. This method was using a dummy gage on a cantilever beam along with the previous single beam with an active gage. They were constructed as half-bridge. These two beams were simultaneously subjected to temperature change. Therefore, the thermal output effects from these two gages canceled each others, and there wasn’t any change in the reading due to temperature changes. The reading is presented in Table 1. Figure 3 shows a half-bridge construction.

Uncorrectable/No Compensation

 In this method, a strain gage on a cantilever beam was wired with two-wire connection. This method of connection is wrong and is not correctable, or there is no way to compensate it.  The strain was affected very largely by the temperature variation. The result from this experiment is also presented in Table 1.

Incorrect Compensation/Correction

Aluminum beam with -0.6-gage

The last experiment was about using a wrong strain gage for a specified material. A strain gage was built as a temperature-compensated strain gage for steel by the manufacturer, but it was used with an aluminum beam in this experiment. There wasn’t any load application, but the strain increased with a drastic range by variation of temperature


The results from Table 1 show that the baseline part of the experiment didn't have any compensation or correction procedure because there is no temperature variation. This part of experiment was performed at room temperature, and the result from this test is considered as actual strain. We need to compare other results with this actual strain.
The simple correction method, which was used to obtain the actual strain due to thermal effect, shows that the result obtained from this method is larger than the actual strain. This means that there is still some source of errors. The strain gage type used in this experiment was unknown, so we picked up a gage factor of 2.055 for this strain gage. Therefore, there was a difference in the actual strain and corrected strain. The vibration of the cantilever beam during experiment can be counted as another factor of causing the error because the testing table was vibrated by some students during performing the test. Another factor is due to initial reading of the strain from P-3500 and expansion of the strain gage itself. The more correction method was used similar to simple method, but the thermal effect was also calculated to the strain gage, i.e. the gage factor was also corrected. The result from this correction method is also similar to the simple method, but it is quite bit smaller in value because the gage factor was corrected. The same errors as simple method were considered to this method.
The result obtain from the self-temperature-compensated, 3-wire strain gage shows that there is a small difference with the actual strain. There is still some effects due to the temperature variation but isn’t that much crucial. The result obtained from the half-bridge, 3-wire construction shows that the readout and actual strains are the same because the thermal output from dummy gage cancels the active one. Therefore, an actual strain can be obtained from direct reading.
The result from a 2-wire connection strain gage shows that there is a large variation due to the thermal output. This method cannot be corrected, or there isn’t any way to compensate it. Therefore, this method is wrong. The result from last part of this experiment shows that there is a large variation in the strain reading even though there wasn’t any load application. The reason was because of selecting incorrect strain gage. This gage was manufactured for a steel substrate, and it was calculated from the steel’s thermal coefficient to be temperature-compensated. This gage was used for an aluminum beam in this experiment. Therefore, the readout was incorrect, and there isn’t any equation to correct it. 


As a conclusion, we can say that the correction and compensation methods in this experiment were reasonable. The errors due to some factors were because of uncertainties in the process of the experiment. The advance correction method was more accurate than simple correction because the correction for gage factor was also calculated. The half-bridge temperature compensation method was more accurate than the self-temperature-compensated strain gage. Two-wire connection of strain gage and improper selection of the strain gage were wrong methods for the thermal output, and there weren’t any way to compensate or correct them. An exaggerated reading was seen for these last two parts. This experiment was an opportunity to get familiar with the methods of error correction and temperature compensation that was caused due to thermal output. In addition, knowing that improper selection of strain gage or incorrect wiring of the strain gage causes a big error.