Post rigs

[Jersey Tom]

Incidentally, a squeak & rattle rig can be used for suspension set-up/assessment, & I would guess that every proving ground has one (at least) of those. All that is required are a set of good transducers (including wheel pan load cells), some conditioning & acquisition hardware & some careful calibration. Oh, & somebody who is familiar with RPC to generate drive files.

Or just get a good driver, and drive over a 'ride road.'

[Belatti]

Incidentally, a squeak & rattle rig can be used for suspension set-up/assessment, & I would guess that every proving ground has one (at least) of those. All that is required are a set of good transducers (including wheel pan load cells), some conditioning & acquisition hardware & some careful calibration. Oh, & somebody who is familiar with RPC to generate drive files.

Whats RCP?

[bill shoe]

Or just get a good driver, and drive over a 'ride road.'

I used to do subjective eval at a PG, including noise. One day I had a bad squeak in the back of a minivan that was driving me crazy, but it only occured while the van was moving. I drove toward the beginning of a ride road, put the transmission in neutral, and coasted while I stooped toward the back to find the squeak. Found it and fixed it, but I only got away with this because the PG had around 1000 acres per driver.

If I had to de-squeak the entire interior of a new model I would appreciate the safety of a shaker rig.

[Jersey Tom]

No kidding. That's cool. The subjective handling and NVH stuff is tough. Think a lot of people have this misconception that test drivers are all about limit handling... which is probably the easiest thing to get down!

[Belatti]

A friend at Uni told me some years ago when he was working in the product engineering department at Ford, he and his experimented boss had to drive arround in a new model to take note of noises, describe them and guess where do they come from.

He took note of 25 noises and he knew where did 8 of them came from.

His boss took note of 135 noises and knew about a large amout of them. Also in some cases the improvement of one of the noises leaded to the supression of another one...

[DaveW]

Whats RCP?

RCP (or RPC, apologies) is an acronym for Remote Parameter Control, an MTS-proprietary suite of software that is designed to compile one or more rig input waveform(s) (drive files) that will minimise the differences between one or more actual & required output waveforms. The software is required for "track simulations", when the required output waveforms are track measurements & all rig suppliers have equivalent capabilities. MTS is (I believe) unique in using the technique to compensate for deficiencies in its own actuators. It is iterative & requires a true guru to make it converge reliably....,

[Dolsum]

Hi! I was just "enjoying" reading 14 pages of post rigs comments, and decided that I should start getting involve!

Thanks Dave and JT to such "rich" discussions and information.

I had some questions regarding the way that a post rig makes the assesment of the cars, as was mentioned that a PI is used and a CPL variation is also used:

- How you calculate CPL variations ? (I do not ask PI as Dave already said somewhere the is a secret!)
- I got some test done on F3 cars, and I see some pitch coupling, pitch damping and heave damping, what do really mean? is it possible to know how you calculate them to get a more deeper understanding?

Great thread, keep it alive!

Cheers,

[GSpeedR]

- How you calculate CPL variations ? (I do not ask PI as Dave already said somewhere the is a secret!)

CPL variation is traditionally an ASD (auto-spectral density) of tire vertical load viewed in the frequency domain, though there are numerous methods that are used in both motorsports and production cars involving manipulations of that data before and after transforming it into the spectral (frequency) domain. Windowing, shaping, filtering are all often used to some degree dependent upon the specimen and test of interest. In the time domain, calculating an Root Mean Square (RMS) Load is also common. There is a significant amount of analysis performed in both the time and frequency domain in addition to the above, but you will start to venture into the proprietary realm...especially if you are interested in calculations of "grip" and it's effects on handling.

Personally, I prefer to analyze in the spectral domain. If you are interested, study up on Fourier/LaPlace Transforms and the properties of Fast Fourier Transforms (FFTs), which you will use with measured (discrete) data. It is easy to get useless/invalid results if you don't keep track of what you're transforming.

- I got some test done on F3 cars, and I see some pitch coupling, pitch damping and heave damping, what do really mean? is it possible to know how you calculate them to get a more deeper understanding?

Great thread, keep it alive!

Cheers,

You will have to be more specific about what you are referring to. I would assume that your pitch coupling is the magnitude of a Transfer Function with a Pitch Accel output and a Heave accel input (likely from a "in-phase" sinesweep excitation), though there are other ways to determine it. Pitch and Heave damping could mean many things depending upon what is instrumented on your vehicle:

1.)Assuming you simply installed your racecar on a rig and attached accelerometers to the hubs and body, heave [pitch] damping could be a measured F-V curve using calculated relative displacements [pitch angles] and CPL load [CPL pitch load].
2.) It could relate to energy dissipated by tires/dampers compared to input energy, where phase angles between actuators and the tires and body become important.
3.) Maybe your F3 had instrumented dampers and shock potentiometers and it is simply a measured F-V curve at the damper.
4.) Perhaps your vehicle was used to generate parameters for a simplified simulation, which is analyzed modally.
5.) ?

The point is that most rigs seem to do things a bit differently; they may even use the same naming and signal convention. You need to direct these questions to the engineers administering the tests, anybody else would be simply guessing (just like me). I'm not aware of any standardization regarding these kind of analyses, so someone please correct me if I'm wrong.

[Dolsum]

Hi!
GSpeeR thanks a lot for taking care of answer me and sorry for not answering earlier, I had been pretty busy!

Regarding pitch and heave, I am not sure how Dave does the measurements, I just get the report of pitch coupling, pitch damping and heave damping, not really sure how he calculate them.

I assume pitch is done by measurement of pitch angle, however no idea how is done the coupling.
I have a couple of new questions! I started a simple Mathcad program (not Matlab as I find it too complicate!) and try to do some "virtual" cases to get a better understanding, and I generate some data in a "lineal" model, so it is easier to get values. I tried to do graphs like Dave does it, for example CPL/Wheel Hub Acceleration in the frequency domain and I have problems as Hub Acceleration sometimes goes through ZERO value, so problems in the division by ZERO! Somebody has an idea how is done!

Thanks to all!

[WilO]

Dolsum,

I believe that the two symmetric modes of a vehicle (pitch and heave), will be coupled in the presence of damping. As such, a pure heave input on the rig will likely elicit some pitch motion. The degree of which would be measured by one of the methods described in GSpeedRs' excellent post.

I shall now get out of the way so that those who actually know what they're doing can contribute.

Wil

[WilO]

GSpeedR,

I was wondering if you might be willing to give us some insight as to how you utilize rig testing; ie. setting spring rates and damper trajectories using heave/pitch inputs (initial setup), and/or track-acquired data 'played back' through the rig to setup for a specific track.

Thanks for your contributions here, very much appreciated.

Wil

[GSpeedR]

GSpeedR,

I was wondering if you might be willing to give us some insight as to how you utilize rig testing; ie. setting spring rates and damper trajectories using heave/pitch inputs (initial setup), and/or track-acquired data 'played back' through the rig to setup for a specific track.

I think this all depends on whether a team has track-derived data acquisition available, as well as the type of racecar and rules it races under.

DaveW described the two predominant methods quite well in this thread, though he obviously has a professional bias towards his method. He may have a better term, but I would describe his methods at MultiMatic as modal characterization. It sounds as though he directly measures modal parameters and then incorporates them into a simplified vehicle model, which is then used to perform modal analysis. It also seems like he performs a lot of traditional (and probably some not-so-traditional) vibration analysis comparing transfer functions between the various degrees of freedom, and calculates energy ratios between components as well. In my experience, this type of analysis becomes significantly more difficult with asymmetry and non-linearity (the latter is an issue for all spectral analysis, BTW, not just characterization). I've never been to the MM rig, so I won't guess as to how springs/bars/dampers are tuned, and I doubt that DaveW would share the cost function that he uses.

Track-playback (using RPC and similar methods) is more popular among teams that can generate track data; rig generated data can be compared directly to track generated data, keeping in mind the limitations of rig testing. Depending on the vehicle, the compromise between body attitude and tire load variation will change from track to track, depending on the strengths and weaknesses of the car and any updates/changes since the last time it was raced. Once a drive file is created, many teams will perform DOEs involving a single (or a few) tunable parameters to determine sensitivities. Certain events around the racetrack will be important and it's up to the race engineers to determine which areas need to be focused on. Analysis of track-playback testing will usually be a more even combination of time and frequency domain analysis, moreso because the average measured values (load, travel, accel) relate to specific parts of the track. Strictly speaking, the drive file generated from track data is only valid for that particular setup...venture far from that setup and you will be seriously pushing the usefulness of your results. Many high-level teams have simulations capable of delivering >90% of the accuracy of this style of testing and thus physical rig testing is only performed to validate their models. I don't believe there are many sims I would trust for specific vehicle characterization, however.

Ideally, teams would implement both methods. Characterization must be done to understand overall vehicle sensitivity. Track-playback is very useful to investigate specific events around the racetrack under more representative conditions. Most importantly, teams need to recognize the limitations of rig testing in general (constrained simulation) to understand what it can and cannot tell you.

All the above is somewhat vague, since the specifics will be proprietary to those running the test rigs.

[Belatti]

completely out of topic, but postrig related after all



Thats me standing over a column. Blue is standing on one leg, brown is "normal" standing, pink is crouching and green is "snowboarding/surfing" position (Im using the postrig to get prepared for the snowboarding season)

[GSpeedR]

Making good use of your active suspension.

[RemcoA]

Hi guys, as a passive reader of this forum, I found myself in a problem in which you might be able to help.

For the sake of virtual seven post rig shaking, I am trying to generate a road surface input for our virtual shaker. We used to use a random white noise signal, but optimizing damper settings on such a road is not useful for racing real circuit.

So we came up with the idea to reverse engineer the damper displacement per wheel into a possible(!) road surface, since actual track data is unknown. I filtered the signal to get rid of displacements due to e.g. cornering and aero. I initially created a Simulink Quarter Car Model, analytically derived the transfer function of the system and the State-Space system. Of course, things like sprung and unsprung mass, spring stiffness, tyre stiffness, damper characteristics are known. However, I did not find a method how to invert the transfer function such that the damper displacement can be used as a "input" to the system, yielding a possible track surface as output. Could anyone help me solving such a reverse engineering problem or give an advise how to tackle it?