Post rigs

[Belatti]

Oh!!! and BTW

What kind of rig test do you think you could do to this vehicle?

[DaveW]

...The phrase "blind leading the blind" comes to mind... but here goes (for an "aero" vehicle):

In 1, to create those "drive files" for the force actuators to simulate deterministic loads, I guess you use any data acquisition system, lets say Pi, right? Wich are the sensors needed? Driver controls, nx, ny, airspeed, 4 x PRL, 4 x upright acceleration, 4 - 6 x damper position, 2 x ride height as a minimum, I think. Temperatures (tyres & dampers) might also be useful.

(I´ll guess and you can correct me)

*Rolling moments caused by lateral acceleration: accelerometer + math function Yes.
*Pitching moments caused by longitudinal acceleration: accelerometer + math function Yes.
*Driven wheel torque reaction: rpm + TPS + math function Yes.
*Aerodynamic pressure: ??? Aero map, airspeed, ride heights (estimated or measured).
*Vertical loads caused by aerodynamic pressure: damper position sensors + math function? No. Aero map, airspeed, ride heights.


Bearing in mind how had we "deterministically" calculated or measured in 1, then in 2, this "complex iterative process" is performed... how extactly? A programmed logarithm? I think here is one of the areas where the expertise of the consultant worths a lot Requires a reply by itself. MTS use "RPC", Instron use "Spidar", not sure what Servotest call theirs. All proprietary. MatLab "mmle3" function & explanation in "State Space Identification" might be helpful. "Consultants" could be useful to advise on what to include in the cost functions used to iterate to a simulation and to assess the effect of set-up changes. Also to advise on what to do when it all blows up (fails to converge), rips a corner off the vehicle, etc.


1) The other forces not represented were the ones some forumers mentioned, such as Fy in tyres? Yes.

2) Correct sensors positioning is critical, right? Absolutely, + scaling.

4) Unless you have created "drive files" for different recorded laps with different lines. Maybe this can be done in order to evaluate if any given setup is more or less forgiven with, lets say, tralibraking to suit some particular driver, etc, etc... Don't know. I was thinking more of driving lines changing as set-up changes.

Im thinking here about a music director: he may know the individual capabilities of each musician but he must make them all sound in harmony and sinchronization.

Returning to race cars, that is: how much my roll center + pitch center variations will let me trailbrake using that spring and damper configuration? A step too far, perhaps.

5) With "more idealized rig tests" you mean simply sending sinusoidal or step inputs to the hydraulic actuators supporting wheel platforms? Yes.

Altogether not totally straightforward. You might consider an alternative approach to be attractive: use hardware-in-the-loop tests to identify & validate an "adequate" mathematical model, & then use the mathematical model to answer more complex questions, accepting that it might not "see" non-linearities like damper cavitation & structural energy dissipation. Either way, there is probably no substitute for "knowing exactly what you are doing" or, to be more precise, knowing what you are not doing.

[marcush.]

haha..the more you know the more you know what you don´t know..springs to my mind.
In effect you still need a guy knowing his trade as it would potentially lead to a disaster just to rely on rig tests as the test alone is no proof things will work
on the track,especially if the rig test does omit -does not account for-some important real life events....

[DaveW]

...as the test alone is no proof things will work
on the track, especially if the rig test does omit -does not account for-some important real life events....

Exactly....

I see several vehicles/year that have clearly been previously set-up on other rigs, some good, some bad, but none ugly. I suspect some of these used track simulations. One customer had set-ups that were consistently over-damped (in my view). After the test he emailed demanding to know why we had "under-damped" his vehicle. We replied that, actually, we didn't think we had, pointing out that the rigid body modal damping ratios returned by his set-ups were lower than ours, despite (actually, because of) our reduced damper settings... A week later, after his next race (which his vehicle won), he did have the grace to email "your stuff worked like a charm", suggesting that what we do works sometimes...

[ubrben]

That's an important point. A car I've attended a rig test with had considerably better damping of the sprung mass pitch and heave modes by reducing the damper forces because the thing was no longer bouncing on it's tyres.

Ben

[Fil]

borrowed from a different thread, originally posted by n smikle..

7-post rig shown & briefly spoken about at 5m20
Interesting to pause at 5m43 for a screenshot of some of the data values Red Bull record.

[youtube]http://www.youtube.com/watch?v=h6RZKqzAaHY[/youtube]

[The_Man]

Let me get this thread back from the dead.

I have spent the entire afternoon read this incredibly informative thread and its tangents. Firstly 'Respect' to DaveW and thank you for taking to time out to explain the way you have.

I think there is one more thing I do not think we have touched upon. That is the manner for the application of the forces on the 7 post rig. As far as I understand the actuators act on 3 points on the chassis. Wherever they are they form a plane and you can simulate pitch, roll, heave or a combination, that is fine. But in reality the loads are actually quite differently applied. For example if we put a formula one car on it the down-force comes from the wings and bottom. When we simulate on a rig we apply forces differently leading to the same down-force, but being applied at different points. So I do not how much this shall effect the results, but I can see the load path information to be quite different to the reality. This came to mind when someone mentioned using strain gauges on linkages to measure the forces. Even during calculation of chassis flex. I imagine the point of application of load is really important.

My question is that is there something done to account for this? Or is this what the 8 post rig does? Does the extra actuator redistribute the loading in a more realistic manner?

One more thing that caught my attention was when someone mentioned the chassis being actuated by flexible cables. Is this true? In the video's they look like cables. In this case how is a situation where the actuator is "pushing" simulated? For example in the extreme case where the inside wheel is almost about to lift in due to roll. Why are not linkages used as actuators, that are capable of pull as well as push? I believe that the actuators are displacement driven rather than force driven but still shouldn't they act in both the ways?

[Sean H]

there are more comlex 11 post rigs out there being used as well.

[The_Man]

there are more comlex 11 post rigs out there being used as well.

I think these are just 7 post rigs that have sideways actuators at the wheels(Hence the +4). The 11 post rig is just a wicked marketing ploy .

On a serious note this is not the first time I hear of 11 post rigs. I know McLaren where talking about one in development about 8-9 months ago. I do not have more information on this though. What else could you possibly actuate?

Steering?
May be the sideways effect of the shark fin engine cover.

[Mystery Steve]

As far as I understand the actuators act on 3 points on the chassis. Wherever they are they form a plane and you can simulate pitch, roll, heave or a combination, that is fine. But in reality the loads are actually quite differently applied. For example if we put a formula one car on it the down-force comes from the wings and bottom. When we simulate on a rig we apply forces differently leading to the same down-force, but being applied at different points. So I do not how much this shall effect the results, but I can see the load path information to be quite different to the reality.

Don't forget that a collection of forces on a body can be combined into a triad of forces and moments about the center of gravity. You can then equate this collection of forces and moments into three forces placed strategically on the body. Typically the loads are applied (on an open-wheeler) close to the nose cone and the rear wing, presumably to make the loads somewhat realistic. I'm not sure how the loading is defined quantitatively. It could be done any number of ways, but probably some combination of data taken from the track, wind tunnel, and CFD.

[Sean H]

there are more comlex 11 post rigs out there being used as well.

I think these are just 7 post rigs that have sideways actuators at the wheels(Hence the +4). The 11 post rig is just a wicked marketing ploy .

On a serious note this is not the first time I hear of 11 post rigs. I know McLaren where talking about one in development about 8-9 months ago. I do not have more information on this though. What else could you possibly actuate?

Steering?
May be the sideways effect of the shark fin engine cover.

a 4 post is just a shaker table, 4 wheels.
a 7 post is shaker table with 3 actuators
a 11 post is a shaker table with 7 actuators (usually each corner plus the previous 3). and yes, some can be horizontal if need be.

[DaveW]

8-post rigs exist, & I have seen one 11 post rig (now deceased). Somebody has an 18 (I recall) post rig.

For 7/8 & 11 post rigs, all but the four wheel actuators are attached to the sprung mass & are "load controlled". It is common (& sensible) practice to attach load-controlled actuators to the vehicle with high structural integrity to avoid instabilities that will often be terminal. In that respect, it could be argued that each actuator moves the subject vehicle further away from being a "runner".

Three (vertical) load-controlled actuators are relatively simple (the proverbial milk-maid's stool). Four (all vertical) are a theoretical improvement (as has been suggested above) but can present control problems (actuators can "fight" one another). The 11 post rig I have seen had three vertical & four horizontal load-controlled actuators (two lateral & two longitudinal), all (to repeat) attached to the sprung mass. An 11 post rig cannot test a vehicle fitted with tyres (sensibly).

In my (slightly cynical) view, an 11 post rig is simply testing the most predictable part of what is left of the vehicle hardware, since most of the unpredictable "bits" (arguably, the reason for executing a "hardware in the loop" test) have been removed (& must, therefore, be simulated).

[Sean H]

a quality shaker table will almost let you put actuators anywhere you need. In the stock car setups, its not uncommon to use more than the 7 total since they deal with more body roll along with aero than open wheel setups.

[DaveW]

a quality shaker table will almost let you put actuators anywhere you need. In the stock car setups, its not uncommon to use more than the 7 total since they deal with more body roll along with aero than open wheel setups.

... & the chassis may be more compliant. Either that or the force-controlled actuators are not attached to true bulkheads. Few road vehicles have bulkheads, & NASCAR's are little better....

[The_Man]

Don't forget that a collection of forces on a body can be combined into a triad of forces and moments about the center of gravity. You can then equate this collection of forces and moments into three forces placed strategically on the body.

I am quite aware of this, but these 'ideas' are strictly for rigid bodies. If one of the aims of the rigs is to check for effects of compliance, then this might not be very accurate. However on second thought, may be the estimate is not too bad away from the points of application of the load.