Creating perception of integrity in auto unibody...DIY methods? 1

[JamesMcD]

It seems like with each new model of car it is claimed to be some measure (30-40%) "stiffer" than the previous generation, laterally or longitudinally. For some cars, there are aftermarket parts to increase stiffness of the unibody, such as subframe braces and strut/shock tower bars. You can create your own as well, simple enough.

My question is this:
How would someone like me, who appreciates and desires "car stiffness", go about measuring vibration and noise in various places in the unibody? And furthermore how would I damp it out?

I am interested in both creating a perception of solidity, as well as actually increasing it. I do not want to use sound deadening products like that heavy tar paper the stereo guys like. I would rather go about it by reinforcing the structure to either make it stronger or change it's natural frequency(?)

You can tell I'm not entirely sure how to pose my question. I hope someone can give me some direction,

thanks.

-James

 

[GregLocock]

I honestly doubt that you will gain much except knowledge and experience by DIY analysis of the structure, but those aren't bad things.

We measure noise with microphones. Asssuming you are trying to do this on the cheap then a $10 microphone is a good place to start. Calibrate it by generating white noise or tones with you PC (this is not a quick project and may involve you in all sorts of reading and development that you don't necessarily want to do).

Measuring vibration for a sensible cost is a bit harder. Typically for human response type problems you only want the frequency range 1-100 Hz, of which 4-32 Hz is the most important, practically. You could try building an accelerometer from a loudspeaker if you get rid of most of the cone, and just leave enough to hold the coil square against the magnet.

An alternative might be to use a straingauged cantilever with a mass on the end, again more research required.

Having got you transducers sorted out then you need to decide where to put them. We typically measure vibration at the seat mounting bolts, the drivers footrest, and the steering column. On a convertible there is a good argument that the header rail should also be measured, as the stiffness perception is strongly affected by the way that it bounces around.

We measure noise at each ear, but I think that is unimportant for stiffness type problems.

You also need to decide how to excite the car. If you use a road surface it has to be repeatable, and should excite the sort of problem you are trying to cure (that may or may not be obvious).

Does that help?
Cheers

Greg Locock

 

[Hatch]

From my perspective, a door which closes with a solid clunk rather than a tinny metallic clang has more quality. That is usually accomplished with either mass (a heavy strong door) or a thin door with damping (tar paper, composite damping sheets etc). A stiff structure will not do much.

Inside the car, a quiet car with a lack of engine noise, squeaks etc is "quality". Some people like to hear the engine, exhaust, etc (Porsche anyone?) but that is not my definition of quality (maybe my age is showing here). A stiff body structure will help with the squeaks.

Back in the days when I lowered my car, used front and rear strut tower braces etc, the reason was for handling. My car behaved like a go-kart and some bumps were painful.

The point of my discussion is that you should define exactly what you mean by "quality" and then try to measure that variable. It might be noise, handling or something else. As you can see from Gregs answer above, vibration and noise measurements are not simple or cheap (the setups in the automotive industry can be quite complex and expensive).

Hope this gives you some ideas.

C. Hugh (

http://www.Hatch.ca)

 

[D2]

Jim,

Here's a strange idea, ive come across - 'Structural Foam'. Foam the cavity in the unibody frame. Never tried it, but its worth a thought.

 

[GregLocock]

Bin there done that. It has two uses (1) it stops acoustic transmission along the members. It is very good for that (2) local stiffening of overly large tubes. It is very good for that. In a steel structure with normal size members I doubt it will help much. In a grp body with locally poor properties eg over the front wheel arch then it has a big (measurable) effect.

I agree with hatch, you really need to decide what you want to fix, rather than just globally stiffening everything.

A good example used to be the whale tail Porsche 911 turbo convertible. The body stiffness was b. awful, the worst I ever measured. Yet the perceived quality wasn't too bad, as they had carefully tuned the shocks, and done a lot of work on clearances, and squeaks and rattles.

Cheers

Greg Locock