The measurement wizard allows to configure your on-line FRF measurement. Note that most settings can be changed during the measurement and also after completion when working off-line. Such changes include:
The following sections describe the five steps in the measurement wizard.
In the first step, a connection can be established with a data acquisition system (DAQ). DAQ hardware requirements are listed in 1.3 System requirements. The PAK MKII hardware can be connected either directly or through the PAK cloud IO service. For typical applications, the direct method is recommended. The cloud method is required when using multiple MKIIs. Using either method, first ensure the MKII is connected to a computer with an ethernet cable and powered on.
The connect hardware step
DIRAC can only be used for measurements with the MKII running in PAK live mode. If the MKII is not in live mode, the firmware can be updated using PAK tunes. Find your device in the “PAK MKII devices” list on the right, and then click “Update PAK MKII” within the Firmware and Licenses section. This may take up to 20 minutes.
To connect directly, enter the DAQ IP address of the MKII (found on the MKII screen) within the Connect Hardware page of the Test Settings card and click Connect. A typical address would be 192.168.100.5. The DAQ status should turn green and show “Running”.
The following video shows how to connect to an MKII directly and gives some troubleshooting tips.
Using the cloud, one computer/server can host the PAK cloud IO service for several measurement devices. Using the host computer with the Cloud IO Service installed, access the service by entering localhost:17671 in a browser. Here, you can add and connect to several measurement devices. It also allows to pull the license (CS_IO_USE) for high data throughput, which is typically required in a multiple-DAQ set-up.
Next, DIRAC will connect to the cloud server. In DIRAC, within the Connect Hardware page of the Test Settings card, enter the IP-address of the computer hosting the cloud service and click Connect. If you are hosting the service on the same computer as you are running DIRAC, enter 127.0.0.1. The DAQ status should turn green and show “Running”.
In case you want to use multiple frontends, it is advised to connect them via a network switch. Make sure the frontends are in the same IP-address subnet, so for instance 192.168.100.xxx.
The following video explains how to connect to an MKII through the cloud.
If more than one system is connected through the cloud, you must enable PTP synchronization. Use the switch ‘PTP synchronization’ to enable this and press PUSH. Beware: this will disconnect DIRAC from the cloud and will reboot the PAK systems for which the state has changed.
The PAK systems table when connected
When the systems are online again, reconnect and observe that the PTP synchronization is indeed “Synchronized”. This might take within a minute to update.
The second step is about the impact measurement settings. Here you can define one or more input sources (impact hammers), set the sampling rate, sampling time, frequency settings and define filtering curves in case more than one input source is defined.
The measurement settings step for a single impact hammer setup
You can add input sources by clicking on the plus icon on the top right. This will open the sensor database under the tab for impact hammers. For conventional FRF testing, it is commonplace to use a single impact hammer only. The pre-trigger settings of the source can be set using a pre-trigger percentage (a percentage of the time window) and a trigger level. In the single-input case, filtering shall always be set to none (more about this further on).
Here the sampling settings are defined. The sampling rate can be set for a list of possible sampling rates in Hz supported by the DAQ. The sampling time defines the length of the measurement window in seconds and can be set freely. The resulting block size is shown below the two values and is calculated from the sampling rate multiplied by the sampling time.
Note that sampling settings cannot be changed when measurement data already exists within the project. If needed, all data can be deleted by right-clicking within the Impact Locations card and selecting “Delete All Measurement Data in This Project”.
The Nyquist factor offers three options: 2.0, 2.56 and Custom. This specifies the effective frequency bandwidth of the measurement. When set to 2.0 or 2.56, the frequency limit becomes the sampling rate divided by that factor. Factor 2.0 is the maximum observable frequency according to the Nyquist–Shannon sampling theorem, whereas the 2.56 is the “conservative” factor commonly used in NVH engineering. For the case of sampling at 16384 Hz, this yields respectively 8192 or 6400 Hz bandwidth, which will be displayed in the field Frequency limit.
The Custom setting allows to manually specify a frequency limit to match the frequency range of interest. Oftentimes this is considerably lower than the sampling rate. Setting a lower frequency limit during testing has the added benefit that all other frequency-domain calculations are performed faster. The option is also non-destructive, meaning that the saved measurement data is unaffected and one may always choose a higher frequency limit later on for analysis.
DIRAC offers the unique possibility to measure with multiple hammers connected at the same time. Just like a good set of audio speakers combines subwoofers, mid-range speakers and tweeters to sound well in all frequencies, so do impact hammers when it comes to exciting a structure. This means that to obtain an FRF function that is well measured in all frequencies, their results need to be somehow merged. Merging of FRFs is sometimes done by cross-fading in a post-processing step. This is known to be tedious and cumbersome, as one needs to find cross-fading frequencies that work for all FRFs.
DIRAC takes a different approach, by implementing spectral filtering up-front in the calculation. This will ensure that each input source adds energy in the frequency range where it is most effective. Also, there is no need to match the cut-off frequency between the input sources; one merely needs to focus on selecting the bandwidth in which that particular impact source is most effective, i.e. renders sufficient signal-to-noise and is free from non-physical response effects. Subsequent merging will occur based on the energy contributions in the resulting APS and CPS.
Filter settings for the input sources with their respective filter shapes.
The input and output spectra that are fed into the FRF calculation are filtered by shape functions that can be defined per input source. Three filter shapes are available: low pass, high pass and band pass. An example is shown in the figure above. The filters can be configured using:
Indeed, the contribution of an input spectrum outside half of the roll-off frequency will be set to zero. If it happens that, as a result of this filtering, a certain frequency range is not captured by any input source, a warning will appear.
Lastly, all settings regarding filtering can be changed afterwards, so won’t affect any recorded data in a destructive way.
Sensors and excitation sources are mapped by dragging and dropping them onto the desired channels on the DAQ system. Multiple channels can be selected and mapped simultaneously. DIRAC checks for compatible modules and channel types, to make sure ICP and Microphone channels cannot be mixed up.
Map sensors step
Some general things to note:
The sensor settings provide configuration options for the measurement channels. The calibration values, mode (ICP/voltage/microphone), voltage range, coupling (AC/DC), and grounding (differential/grounded) of the sensors can be set here. Other information about the sensors. e.g. name and serial number, were previously defined in the Prepare Module and must be updated there. By clicking “Show advanced”, high-pass and low-pass filtering can be defined.
The table will show the values as they are configured in DIRAC, which is information that gets stored in the project. If an inconsistency is observed between the channel definitions in DIRAC and the actual values on the DAQ, this will be marked by a yellow background on the value field. Here, the user can choose from two options:
Sensor settings step
For TEDS-enabled sensors, the TEDS information can be read out by clicking the TEDS button next to a channel, or all TEDS information can be read out simultaneously using the button in the lower right.
TEDS is the memory within some sensors that is utilized for storing information about the sensor, such as manufacturer name, sensor type, model number, serial number, and calibration data.
If the TEDS button is pressed for a sensor that does not have TEDS, old sensor information from PAK may be retrieved that is not relevant for that sensor and the sensitivity may be updated incorrectly. The TEDS readout button should only be used with sensors that definitely have TEDS and verified afterwards.
The last step is about the impact windows that are used in the FRF calculation. This is shown in the figure below. The parameters for the force/exponential window can either be entered manually on the right, or the orange bars within the plots can be dragged to the desired parameters. The most recent impact is shown in the plots.
Once the last step has been completed, a summary page shows the settings as set, alongside an encouraging image to get started measuring! Now would be a good time to save the DIRAC project though, for instance by pressing CTRL+S.
Contact us for more VIBES
Contact our support team or call us on +31 85 744 09 70