NVH questions

Ask questions relating specifically to the PicoDiagnostics NVH equipment and software here.
Post Reply
DS4000
Newbie
Posts: 3
Joined: Tue Oct 27, 2020 4:45 pm

NVH questions

Post by DS4000 »

Good afternoon,

I have some questions on the PICO NVH software,

1 - Can the software playback the captured NVH signals as an audio signal ? I appreciate that there is a signal generator to use, but thinking if you could playback the fft converted samples as an audio file - it may help with particular fault finding exciting resonances in the car structure etc ...

Part B of this question... ( :D ) a car audio system may go down to 50Hz as an example, but quite a few NVH frequencies of interest may be alot lower - say T1 at approx 12Hz. Also using the PICO signal generator, is there another drive method of creating these lower frequencies ( eg: solenoid ) rather than using a cars audio system ?

2 - I am investigating a T3 / P1 vibration. ( Unfortunately the rear drive ratio is 2.86, so T3 / P1 very close ). Can the resolution in the PICO nvh software accurately distinguish between a T3 / P1 in this case ?

3 - If there was a tyre/axle vibration recognised, in order to narrow down which tyre was the culprit - I presume moving the magnetic accelerometer to each corner of the car and repeating the road test would allow you to isolate which was the problem area ? Or does a car body structure distribute vibrations to such a degree that this is not possible ?

Full disclosure - I am currently using the J.Kelly NVH app and while I think it is excellent for what it is - I have come to a point where I need more granularity over the results - hence my above questions to see if the Pico NVH is the next best route for me.

Many thanks.

Steve Smith
Pico Staff Member
Pico Staff Member
Posts: 1578
Joined: Sun Aug 25, 2013 7:22 am

Re: NVH questions

Post by Steve Smith »

Hello and thank you for the post

1. You can playback microphone data immediately within the NVH software (post capture)
Alternatively, you can highlight the recorded data in the signal history and export in .wav format where you can listen to playback using 3rd party software/media player
The following link will help topic21639.html

Low frequency creation: I had not considered ultra-low frequency generation (12 Hz) via the In-Car Audio System and so will revisit the effectiveness of such a test.

Could I ask, would you be recreating such a low frequency vibration to simulate the effects upon the cabin from T1?

When you mention “Solenoid” as an example, how would you apply this device to the cabin or would the solenoid be used to apply a vibrating force to a component? (Forgive any misinterpretation of your question)

2. Regarding T3 and P1 with a 2.86 differential ratio we have the scenario that these frequencies are very close together within the spectrum

For example:
T1 at 10 Hz exhibiting 3 disturbances for one revolution of the tyre = an offending frequency of 10 x 3 = 30 Hz
P1 frequency = T1 x 2.86 = 10 Hz x 2.86 = 28.6 Hz which is very close to T3 at 30 Hz

To separate these frequencies, you can adjust the Response and Accuracy settings of the FFT.
The following link will help viewtopic.php?p=82781#p82781

With that said, the existing settings should provide adequate resolution for you to define T3 form P1 using the Frequency view.

Depending on the vehicle and drive train arrangement it may be possible to deactivate the rear propshaft for AWD vehicles when using an Economy mode (Food for thought)

Applying load to the drive train in terms of acceleration and deceleration may also help to pinpoint either T3 or P1 as your offender.

Shaft’s and joints are dramatically influenced by load, whereas T3 (caused by tyre deformation) is generally present regardless of load.

3. Pinpointing vibration (or zoning) is most certainly carried out by relocating the accelerometer as you describe. Ideally using multiple accelerometers, keeping one accelerometer on the driver’s seat bolt (as your cabin reference) and others spread about the vehicle

The following video includes an animation of this technique https://www.youtube.com/watch?v=gWI4h4r ... Automotive

Please note there are 7 videos in the series and each one is well worth a visit in order to familiarize yourself with our NVH solution

You mentioned the influence of the body structure upon the zoning of vibrations; this is most certainly a variable we have to deal with when using an accelerometer as a stethoscope.

Ideally, when comparing vibration amplitudes from one side of the vehicle to another, ensure the accelerometer is connected to a similar mass

Suspension and subframe bolts make for great mounting points as they are generally of a greater mass, mirror imaged on the vehicle (left and right) and secured beyond reasonable doubt to the chassis

In this scenario, the measured vibration amplitudes will be comparable rather than mounting the accelerometer on a suspension bolt on the left side of the vehicle and then comparing amplitudes to an accelerometer mounted on a flexing body panel on the right.

I hope this helps, take care…..Steve

DS4000
Newbie
Posts: 3
Joined: Tue Oct 27, 2020 4:45 pm

Re: NVH questions

Post by DS4000 »

Hi Steve, thanks very much for your reply - that answered alot of questions.

Regards to the audio response in a car, I believe your average car audio system will not reproduce our fundamental frequencies of interest T1,E1,P1 etc ... very well at all. I did find that there are things called 'bass shakers' which effectively speakers(without a cone) with a weight attached to the voice coil. These operate as low as 5Hz and they are mechanically attached to a structure to transfer the vibration. Used in home cinema etc ...
( This is what I was referring to as a solenoid )

Since I was thinking that with your signal generator function, it would be a great teaching aid to actually use something like the above and output a frequency that equated to T1 at 50mph etc ... Ie this is what a T1,P1 etc ... feels like.

Thats very interesting that with the NVH software you can highlight portions of the accelerometer data and save as .wav file - this was exactly what I was after. So doing this - you could actually replay your 'test drive' frequencies of interest within the car environment.

Even go so far - as bolting the 'bass shaker' to say a suspension component and see how much of a transfer of this vibration you then get through subsequent parts of the car. It feels like it might give an extra dimension in fault finding road borne vibrations back in the workshop. I'm just thinking out loud with this one, since I think there are lots of variables, like vibration amplitude and if the x,y,z plane of vibration is critical in highlighting the problem source etc ...

Nick.

Steve Smith
Pico Staff Member
Pico Staff Member
Posts: 1578
Joined: Sun Aug 25, 2013 7:22 am

Re: NVH questions

Post by Steve Smith »

Hi Nick, thank you so much for the feedback

"Necessity is the mother of invention" and your suggestion of a vibrating plate/shaker attached to a suspension structure is something that has bugged me ever since this case study topic21962.html

Long story short, to be able to introduce a vibration of 111 Hz (in the case above) to various suspension components whilst stationary in the workshop would be a joy.

There are shaker plates available on the market but I have no doubt the cost would run into many £££££££'s

Please keep thinking out loud Nick and share your thoughts here

Thank you again for posting, take care......Steve

DS4000
Newbie
Posts: 3
Joined: Tue Oct 27, 2020 4:45 pm

Re: NVH questions

Post by DS4000 »

So a little update, I was trying to apply a particular frequency to individual components to help in my fault finding.

In actual fact I have, to a point ...

Initially I tried a low cost cinema bass shaker - and the main problem was the extremely non linear response ( even away from its resonant point ). A further problem with this was mounting the assembly to any part of the car.

My next solution was to use a eccentric weight vibration motor. By knowing its rpm, you can calculate the applied frequency and using a PWM controller can vary the rpm and hence the applied frequency.
The motor was mounted in a DIY sprung clamp which made it very easy to apply to any car component.

Did it work ? well yes, surprisingly and to a point where you could clamp it to a front wheel and measure it on the steering wheel in the car. So the drawback....... Any ideas ?

.......................................................................

The rpm was varied by the PWM controller fine BUT the acceleration force measured was dependant on the rpm. Ie , lower the frequency, lower the force. Higher the frequency, higher the force etc ....

Thinking back to school days, F = mass X acceleration - so I'm imagining that while weight is constant the acceleration of the mass is changing dependant on the speed of the motor.

I'm imagining this technique is very niche, but I guess you could actually have a calibration curve to obtain a slightly more linear response to a sweep.

Steve Smith
Pico Staff Member
Pico Staff Member
Posts: 1578
Joined: Sun Aug 25, 2013 7:22 am

Re: NVH questions

Post by Steve Smith »

Thank you so much for taking the time to feedback and share in your experiment

I can see the drawback immediately if the eccentric weight is fixed

If we wish to apply a force of 30 mg at a frequency of 10 Hz we may not be able to achieve such an acceleration without increasing the frequency of the motor

Increasing the frequency of the motor may well achieve an acceleration of 30 mg but not at our desired frequency of 10 Hz (Thinking of T1 here)

Does the vibration motor allow for the addition of weight or is this a fixed calibrated unit?

This all sounds very intriguing indeed

Thank you again for the feedback, take care....Steve

Post Reply