Want to use an impact hammer to excite the full frequency range of a structure? Multiple modal hammers, each spanning a different frequency range, can be combined to get a full frequency response of a test structure as illustrated in Figure 1.
Figure 1: A large modal impact hammer (green) produces a Frequency Response Function (FRF) with valid lower frequency content. A small modal impact hammer (red) produces a Frequency Response Function (FRF) with valid higher frequency content. Together the two hammers produce a FRF that is valid over the combined frequency range (bottom).
How does this work? Different hammer tips get better results in certain frequency ranges. For example:
A large modal impact hammer with rubber tip will tend to excite low frequencies.
A small hammer with metal tip is good for results at high frequencies.
Combining the low frequency response and high frequency response of the two hammers (measured at the same location) yields a single Frequency Response Function (FRF) that covers a larger frequency range than a single hammer can measure.
Simcenter Testlab Neo Impact Acquisition software allows a user to set up and use multiple hammers on a structure in the same measurement campaign, which eliminates the need to retest with different equipment.
This article explains how to use the multiple hammer approach in Simcenter Testlab Neo Impact Acquisition.
Contents: 1. Why Use Multiple Impact Hammers? 2. Simcenter Testlab Neo Dual Hammer Example 3. Multiple Hammers in Simcenter Testlab Neo Impact Acquisition 3.1. Channels 3.2. Hammer Setup 3.2.1 Hammer Settings Setup 3.2.2 FRF Merge Filter Setup 3.2.3 Setting up a Campaign 3.3 Measure 4. Simcenter Testlab
1. Why Use Multiple Impact Hammers?
There are different qualities of impact hammers that can result in exciting different frequency ranges of Frequency Response Functions (FRFs). For example:
Hammers with a large mass excite heavy structures better than small hammers by inputting a high level of force.
Hammers with a small mass input lower amplitude force levels and excite a wider frequency range than hammers with a larger mass.
Varying the stiffness (rubber, metal, plastic, nylon,…) of impact hammer tips changes the frequencies produced by an impact hammer hit.
The mass of a hammer and the stiffness of the tip combine to dictate the frequency range excited by the hammer. The longer the hammer is in contact with the test structure, the lower the frequency range produced by the hammer hit (Figure 2) and vice versa.
Figure 2: Example of the difference between a rubber tip (yellow) and a metal tip (magenta) in the time and frequency domain. A short event in the time domain (left, magenta) creates a broad frequency response (right, magenta), while a long event in time (left, yellow) results in a narrower frequency range (right, yellow).
When performing a Fourier Transform on a signal, short duration events in the time domain have broad response in the frequency domain and vice versa.
Some examples of how the mass and stiffness of the hammer affects the duration of the impact:
A soft tip (for example: rubber) causes the hammer tip to stay in contact longer with the test object than a stiffer tip (for example: nylon or metal). The resulting force excitation is focused over a lower frequency range.
Conversely, a hard tip (for example: nylon or metal) produces force excitation over a wider frequency range than a softer tip (for example: rubber).
Due to inertia, the higher the mass of the hammer head, the longer it stays in contact with the test structure. The longer it stays in contact, the lower the frequency range. Vice versa is also true.
Historically, a test would have to be run two separate times with different hammers to ensure the full frequency range of interest was well measured. This is time consuming but also results in two distinct FRFs.
In Simcenter Testlab Neo Impact Acquisition 2406 and above, a test can be performed with two hammers at the same time, and the FRFs are automatically merged to get the performance over the entire frequency range of interest.
The low frequency large hammer and high frequency small hammer both impact at the same location, and the results are measured at a common accelerometer response location. The low frequency portion of the large hammer FRF is combined with the high frequency of the small hammer FRF to get a merged FRF of the best results.
In the example below, two hammers were used to test a metal car frame (Figure 3).
Figure 3: A triaxial accelerometer on a car frame (left) and two hammers used to test (right). One hammer is large with a higher exciting mass and a rubber tip, and one hammer is smaller with a metal tip.
A large hammer with a rubber tip was chosen to put more energy into the low frequency components, and a small hammer with a metal tip was chosen to excite a higher frequency range. In the setup stage, the large hammer was found to be better at exciting low frequencies from 0-170 Hz, and the metal hammer was better at exciting frequencies from 170-500 Hz.
After testing, a coherence plot, which is an indicator of the quality of the measured Frequency Response Function, was calculated and plotted below for the individual hammers (orange and green) and for the combination of the hammers (blue) as shown in Figure 4 below.
Figure 4: Top - Coherence of car frame test for using a small hammer (orange) and a large hammer (green). The large hammer has a cleaner and better coherence close to one for frequencies below 170 Hz. For frequencies above 170 Hz, the large hammer becomes noisy, and the small hammer is better. Bottom – Coherence function for the combined FRF (blue) of the large and small hammer.
Coherence can have a value between zero and one as a function of frequency. When repeating the FRF measurement, zero indicates no repeatability in the FRF measurement, while one indicates perfect repeatability. If an adequate amount of force is used to excite the structure, the coherence at the corresponding frequencies will be close to one.
Looking at the coherence plot, the small hammer by itself does a poor job of exciting the lower frequencies. The coherence plot is noisy and not close to an amplitude value of one. The large hammer, however, has a clean straight coherence at these lower frequencies until about 170 Hz. At this point, the large hammer becomes noisy, and the coherence is poor, and the small hammer gives a better frequency response function.
Using the multiple hammer feature of Simcenter Testlab Neo results in a single FRF and coherence measurement that is well excited over the whole frequency range of interest. The exact steps for performing a multiple hammer measurement in Simcenter Testlab Neo Impact Acquisition is covered in the next section.
3. Multiple Hammers in Simcenter Testlab Neo Impact Acquisition
Simcenter Testlab Neo Impact Acquisition supports using multiple hammers to create a combined FRF spanning over the entire frequency range of the hammers used in the test.
3.1 Channels
Begin by connecting impact hammers to the data acquisition system. Open the Simcenter Testlab Neo Impact Acquisition application and go to “Channels” tab under “Instrumentation” as shown in Figure 5.
Figure 5: After starting Simcenter Testlab Neo Impact Acquisition, go to the “Channels” tab in “Instrumentation”.
In the channel list/grid, more than one reference channel (which corresponds to a hammer) can be turned on at the same time. In Simcenter Testlab “Classic”, it was only possible to select one reference channel (Figure 6):
Figure 6: Important parameters in the channel setup for properly setting up multiple hammers.
In the channel grid, set the following:
ON and Reference (Orange): Each hammer channel must be turned on, and the reference box must be selected for every hammer. If this does not happen, the hammers will not be able to be set up later.
Point ID and Direction (Green): Enter appropriate name and direction for where the hammer is located. Keep in mind that any identification entered here can be overwritten by the Campaign Setup later.
Measured Quantity (Magenta): Ensure the measured quantity is set to force for all of the hammer input channels.
After setting up the hammer and accelerometer channels, navigate to the “Impact” task at the bottom of the screen and click on “Hammer Setup” subtask as shown in Figure 7.
Figure 7: “Hammer Setup” in the Impact tab.
The next sections explain the setup of multiple hammers in further detail.
3.2.1. Hammer Settings Setup
Under the “Hammer Settings” menu in the upper left of the screen, click on the drop-down menu next to “Reference channel” to select a hammer to assign. If the “More” button is clicked, additional properties of the hammer can be set, such as the target force level, the type of tip used, and the hammer mass (Figure 8).
Figure 9: Selecting a hammer channel and typing in the hammer properties.
The more button contains annotation properties of the hammer that are useful for keeping track of the hammer properties, but do not have any effect on the results or measurement. If a user is using the same hammer repeatedly and wants to save these properties, these can be done by clicking the save button in the corner menu.
To add an additional hammer, click the add hammer icon, and select the reference channel for that hammer (Figure 10).
Figure 10: Adding additional hammers and naming each one by clicking on the default name and typing a new one.
More than two hammers can be added. To name the hammers, click on the default names in the left top corner of the box and type in the name.
Press the “Start” button in bottom left corner to begin executing practice hits with the hammer. Hit the structure with both hammers a few times to get and idea of the trigger level used. When finished with the practice hits, click the “Stop” button in the bottom left corner.
Setting the trigger level is a manual process in Simcenter Testlab Neo Impact Acquisition. With the practice hits still shown in the scope hammer display, right click in the display and hit “Add Single Cursor -> Y (Front)”. Then take the cursor and align it so that it goes cleanly through all the impacts as shown in Figure 11:
Figure 11: Adding a Y-cursor (right click and “Add Single Cursor -> Y (Front)” to the scope hammer display to find the trigger level amplitude.
Note the amplitude value of the cursor and type it into the trigger level box for the hammer as shown in Figure 12:
Figure 12: Using the cursor amplitude to type in the trigger level in the dialog box for each hammer.
Repeat this process for each hammer. With both hammers setup, each hammer will be automatically recognized in the measurement phase.
3.2.2. FRF Merge Filter Setup
Next setup how the merge will be performed for the FRFs of the two different hammers. Click on the “Multi-Hammer Filter” button in the hammer settings icons as seen in Figure 13 below:
Figure 13: The multi-hammer filter button.
The following is an example of how two different hammers can be configured based on frequency excitation (Figure 14).
Figure 14: Setting up the hammer filters for multiple hammers based on the frequency excitation of each hammer.
For this dual hammer setup, “Low-pass” and “High-pass” settings will be used with each hammer. As seen in the data, 170 Hz is the point where the large hammer is no longer beneficial, and the small hammer excitation becomes more consistent:
Low-pass: Based on the spectrum excitation from the practice hits, the large hammer in this example works better until 170 Hz. Since the large hammer has a higher amplitude in the lower frequency range, a low pass filter option will be selected from the drop-down menu and the “High-end center frequency” will be set to 170 Hz.
High-pass: For the small hammer, a high-pass frequency filter is applied, and the “Low-end center frequency” set at the exact same frequency of 170 Hz.
These should result in an equal cross-over between the two filtered FRFs. In this example, the merged FRF result will only contain the frequency response of the low hammer up from 0 to 170 Hz, and the small hammer at 170 Hz and beyond.
It is important to note that since these are filters, the transition width needs to be set such that there is no drop out in the spectrum filter display (Figure 15).
Figure 15: Do not do this! Example of a filter where the transition width doesn’t overlap, resulting in a section of frequencies being completely filtered out. In this case, the "low-end" and "high-end" frequencies should match to ensure a even transition.
If the filter goes to zero, it will filter out important data at those frequencies and the FRF will not appear to have any modes or excitation there. The software will warn if this occurs. The frequency cutoffs of the individual filters should match instead.
3.2.3 Campaign Setup
This section will cover important settings in the campaign tab that are required for a multi-hammer setup. In the “Campaign Setup” subtask, click on the “Add Excitation DOF” button and select the node on the geometry where the hammer will be impacting (Figure 16).
Figure 16: Setting up a campaign for an impact location. Both hammers will impact at the same location.
Here the direction and point ID that were initially skipped in the channel setup will automatically be set and renamed. Select the node and direction of impact and then press “OK”.
If the current project does not have a geometry, text can be directly entered into the "Excitation DOF" field instead. If using multiple hammers, enter the same name for both in the "Excitation DOF" field.
The panel on the right side of the screen called “Campaign Settings” has three important settings to check if multiple hammers are being used as shown in Figure 17:
Figure 17: Key settings for multi-hammer use in the "Campaign Settings" menu.
Detailed explanation of the settings:
“Target impacts per hammer”: This setting will tell the software the minimum number of impacts it expects to see from each hammer before it considers a measurement to be done. For example, the default number of target impacts is set to five (figure 14). This means that if ten impacts are done with one hammer, but only four are done with the second hammer, the measurement will not be considered complete until at least both hammers reach five impacts each.
“Impact mode”: If using a roving hammer/excitation method, click “Roving excitation” in the drop-down menu. Multi-hammers are also supported in roving-accelerometer applications.
“Minimum impacts “ for processing needs to be set. The averaged FRF will at least have this many hits included in the average.
"Minimum impacts" relates to how the feature “Smart Hit Selection” will choose hits. To read more about how to use Smart Hit Selection, read here: Smart Hit Selection in Simcenter Testlab Neo.
Move to the “Measure” subtask of the “Impact” task as shown in Figure 18:
Figure 18: The “Measure” subtask under the “Impact” task is used to start measuring.
Click the “Start” button in the bottom left corner to begin measurement. Take impacts and measurements on the structure. A table of hits is filled in as the measurement progresses (Figure 19).
Figure 19: A list of all measurements taken and corresponding hammer is made during the measurement. The software will choose the best hits of multiple hits taken using smart hit selection.
The multi-hammer functionality in Impact Acquisition automatically senses which hammer is being used and does not require all hits to be done in a particular order. A user can swap back and forth between hammer hits as desired.
Smart Hit selection is on by default and will automatically choose the best combination of hits based on the minimum impacts set in the processing settings. These hits will be used to create the merged FRF.
To save the FRF, finish the measurement, and then click the “Save” icon (Figure 20) in the bottom center of the screen.
Figure 20: Saving the FRF results to the project.
Name the results and press OK. These resulting merge FRF should now appear in the project file.
To view the merged FRF, either open Simcenter Testlab Desktop Neo, or go to the “Desktop” task at the bottom of the screen. Click on the active section where data was saved and navigate the folders to find the FRF (Figure 21).
Figure 21: Plotting the final FRF results after measurement in the “Desktop” task.
The final merged FRF and coherence result (blue), overlaid with the data from the separate hammers (red and green), for the vehicle example is shown in Figure 22.
Figure 22: Merged FRF and Coherence results compared to individual hammer results for a car frame subassembly test.
The FRF consists of the large modal hammer data below 170 Hz, and the smaller modal impact hammer data above 170 Hz.
Here the merged result contains the best overall coherence by combining the low frequency response of the large hammer and merging it with the high frequency response of the small hammer.
4. Simcenter Testlab
Don't have access to Simcenter Testlab Neo 2406 and higher? For FRFs that have already been measured, use the "FRF Merge" program located in the Tools menu (Figure 23):
Figure 23: The "FRF Merge" program is used to combine already acquired low frequency and high frequency FRFs.
First start Simcenter Testlab and open the project containing the low frequency and high frequency FRFs to be merged. Then run the "FRF Merge" program.
