Contemplating a Build

Is there a common favorite CFD software among everyone?

OpenFOAM
Solidworks
Fluent


I've tried to open OpenFOAM it won't even come up. Think I'm gonna have to make a new partition for Linux and try again.

Is there anything that ties well into Python for automation scripts? or MATLAB and Simulink?
Simulink would work well to feed back into the vehicle simulator but if not it's just added work which is fine.

Don't expect any CFD printouts yet, this is to get ready. Learn the new software. At present I'm busy working on Dad's christmas present, engineering drawings for his Land Roller at the farm. After which I'm gonna work some more on my suspension points generating computations. I wanna make what I've got more 3D by including tube stiffnesses and use the current geometry stiffness to my fitness function.

Common CFD… if you know what you do you can get more with OpenFoam (that is if you want to develop the code rather than your race car). If you have access to commercial solvers than you should do it. Don’t even consider SolidWorks as a solver. All major suits (CFX/Fluent/Star) will work fine. Don’t expect that advertisement stories about easy setup of optimisation run by coupling different software is how it actually works – in real life things are not that bright. Your main thing (workload) is mesh preparation and that will take the most work. Automating (butch meshing) after you got it right on baseline model will save lots of time when iterating different solutions. Expect some pain (meshing) with complex model such as open wheel car. Think in advance when deciding at which level you defeature your model. Think in advance what (details in your model) will effect the flow and what will effect the solver – act accordingly. When you start understanding the flow than forget about taking advantage of symmetry – you’re interested in what happens at high (bad moments) yaw angles.

Automated PRE (if you keep model/parts IDs the same) will save lots of time.

Automated (batch) postprocessing will help a lot in comparing of many result files.

Etc., etc.


Main thing – don’t expect hard numbers but rather look for tendencies and understanding of flow.

Bad news - for an open wheel race car a 3-4e7 cell mesh is .... starting point. You'll need some computing power.

Cheers
Ted

I'm about to re-code my suspension points generating program and had a couple questions.

What does everyone else think about trying to get the kinetic roll center to move left and right with the force based roll center? My design calculations and ideas while doing SAE hadn't advanced enough for me to consider the idea. But I've progressed in my techniques, and I had an intuition that for my next car I might try to get a roll center that stays a constant height (importance is never understated) however can move left and right following the force based roll center. My intuition is that this could be a decent idea, or should I stick with a completely stationary Roll Center which (on paper anyway) does not move much more than 10 or 20 thousandths of an inch? If kinematic roll center moved opposite to force based roll center the pulling down of the inside wheel would work against the pushing up of the outside wheel in a turn and reduce jacking .... not that the roll centers are very high to begin with but would keep the chassis more stationary in theory.

I thought I would model two four bar linkages, solve for the lines of the UCA and LCA find their instance centers. Solve for the equation of the line from tire center to the instance center on both sides and solve the intersection of that for roll center. In this way I could make a function "at 'y' roll angle of the chassis points, roll center should be 'x' far left and right." Chassis roll being approximately linear and proportional to weight transfer, it should allow me to design a kinetic roll center which moves with or against the force based roll center by a proportion I specify.

I'm currently using the math from a paper done by a man who worked for GM "Suspension Synthesis for an N:1 Roll Center Movement" which gives me the length proportions of UCA and LCA for a suspension with very little roll center height change as the chassis points roll. I'm just thinking of growing what I'm already doing to the next level. But I've never spoken to anyone else who's purposely tried moving their kinetic roll center left and right and what behavior that displayed. It's not in any textbooks or published journals I've read. Since most of my professors work on semi trucks they're not much help. I thought maybe someone on here has tried it and can share their insights.


I also find myself wanting to know more about Pushrod on Upright designs. I know if you place the pushrod ahead of the steering axis you'll reduce the weight transfer across the car due to Mechanical Trail and Caster angle which I would like to exploit and run high caster angles. If you put the Pushrod inboard of the steering axis you lower the front chassis under steering angle and raise the chassis height when steered straight. Someone has mentioned using that for an Aero advantage, more downforce in the corners, but I can't imagine it was that drastic of an advantage.

I'm much more curious about tuning steering feeling through P.O.U. or the trade off between steering feeling and purposeful weight transfer and ride height manipulation. I'd like to ask for driver insights of people who've driven a car with this design and how they were to drive. The feeling of break away when traction is lost in the rear, was it more or less noticeable?
the feeling of loading up on the steering, did it still feel intuitive to the tires? or did you loose intuition of what the tires are telling you?


Thank you in advance for any insights you can provide into the fascinating topic of Pushrod On Upright and purposeful Kinematic Roll Center movement

Roll center location and movement are critical to get a car to handle as you want. I start a car setup with some starting point for roll center location and then I tune the car by moving that point and changing how the roll center moves dynamically. That is the fundamental variable in car setup.

You are trying to chart new territory. I really urge you to get a FF, start driving, and use that car as a platform to develop your ideas.

Alberta, are you on the FSAE forums? I'm pretty sure Z has written a couple novels on roll center migration, that'd be a good place to start.

Yeah I'm familiar with that forum. It's not so obvious on that forum to seperate opinion from published scientific concensus. I read up some from that forum and came away .... I'll just stick with a stationary roll center.

And pushrod on upright needs some simulation time ... there's a few important geometry considerations to account for. Too much you probably can't just make generalizations. Rising rate spring rates or a bit of longitudinal geometry component in bellcrank placement and it all starts adding additional dimensions.

Best to simulink it and see what spits out the other side

Quote:

---

Originally Posted by AlbertaSpeedShop

Yeah I'm familiar with that forum. It's not so obvious on that forum to seperate opinion from published scientific concensus. I read up some from that forum and came away .... I'll just stick with a stationary roll center.

And pushrod on upright needs some simulation time ... there's a few important geometry considerations to account for. Too much you probably can't just make generalizations. Rising rate spring rates or a bit of longitudinal geometry component in bellcrank placement and it all starts adding additional dimensions.

Best to simulink it and see what spits out the other side

---

Pushrod attached to the upright has some merit but it complicates the packaging. The JDR uses this technique with decent success.

When putting the push rod on the front upright, the location is very critical. The effect varies a lot with small changes in position. Many different effects can be achieved.....some are beneficial....some are not. You have to know what you are trying to achieve...and spend a lot of time finding the position on the upright that yields that result.

Jerry Hodges
JDR Race Cars

I will add to what Jerry has said that if you can pull it off correctly you will be very happy with the results.

I did it at the rear on several designs of FF and FCs

So, I would think once you get a simulation working of POU, what you're looking to do is combine the MZ output of both inboard and outboard tire to get to a steering wheel torque function.

Then minimize a reimann sum of the second derivative of steering effort? First derivative would be a function of the tire, but second derivative, the change in how the steering effort loads up, that would be a function of the geometry you specified in the original design yeah? so the lower the sum of the second derivative through the range, the more linear the steering feeling is. But is linear what I want? Seems to me I could also just make a non-linear function of steering effort I desire, and then minimize the error between the effort produced and the function I made up. In-so-doing I can make a steering effort that would feel to the driver as if the tires hit a wall and provide an exponential feedback of whats going on at the tire on the limit. hmmm the possibilities.

Tires squeel when they've defiantly been pushed too far ... so maybe that's all I need for feedback, and a linear steering effort provides the most predictable feeling to the driver.

I suppose I should examine that model on a straight ahead road with a bit of a side slope to it as well, that the feedback is still intuitive when there's weight transfer and roll without steering input

And keep steering effort below 25 ft lbs. That's a value I've seen elsewhere, does that value make sense to everyone here?

Try a KPI of 6 degrees, caster 3 degrees and minimize the scrub radius. You get any serious down force, which a FB will make and you will be at the limit for many drivers.
Down force increases steering effort by a bunch. You can lower the geometric steering forces and then the driver looses feel and you reduce the centering tendency of the car.

Ever heard the expression "over thinking it"?

Quote:

---

Originally Posted by AlbertaSpeedShop

So, I would think once you get a simulation working of POU, what you're looking to do is combine the MZ output of both inboard and outboard tire to get to a steering wheel torque function.

Then minimize a reimann sum of the second derivative of steering effort? First derivative would be a function of the tire, but second derivative, the change in how the steering effort loads up, that would be a function of the geometry you specified in the original design yeah? so the lower the sum of the second derivative through the range, the more linear the steering feeling is. But is linear what I want? Seems to me I could also just make a non-linear function of steering effort I desire, and then minimize the error between the effort produced and the function I made up. In-so-doing I can make a steering effort that would feel to the driver as if the tires hit a wall and provide an exponential feedback of whats going on at the tire on the limit. hmmm the possibilities.

Tires squeel when they've defiantly been pushed too far ... so maybe that's all I need for feedback, and a linear steering effort provides the most predictable feeling to the driver.

I suppose I should examine that model on a straight ahead road with a bit of a side slope to it as well, that the feedback is still intuitive when there's weight transfer and roll without steering input

And keep steering effort below 25 ft lbs. That's a value I've seen elsewhere, does that value make sense to everyone here?

---

As I said.....you need to decide what effect you are looking for. Run a bunch of simulations with different positions on the upright and look at the results. You will see several different effects.....but to choose one over the others, you need to know how each will effect the cars performance.

Jerry

Quote:

---

Originally Posted by S Lathrop

Ever heard the expression "over thinking it"?

---

This is my favorite saying, I use it daily on my employees.

2. Dampers must be independent from each other with no interconnectivity.
However, data acquisition is permissible, as long as it serves no
other purpose.

Could someone help me with where the line is drawn in the sand on this point?

If I had a damper connected so that it's compressed when FL and/or FR is in bump this would handle the ride spring rate for front axle. and another damper which is compressed by FL but extended by FR when both wheels move in the same direction (or vis-versa), this damper would handle roll across the front axle. This would also be fine correct? I'd use 4 dampers each damper is independent of the other correct? or are the bellcranks considered to interconnect them?

Is this okay? Or is the fact that one side is connected to left wheel and the other side connected to right wheel at issue? Traditional systems are fine because it's an undamped spring being connected. Where as an arrangement like this is banned because it's across a damper?

Some years ago, Penske had a 3rd spring setup where the two shocks were interconnected hydraulically. Both shocks fed oil to a third cylinder and valving system. Not allowed today.

The other reference to data logging is to prohibit any form of active suspension. I would interpret that to include any system that changed shock function electronically and tied to the data system of any type.

You can use as many shocks as you want. The shocks can be interconnected mechanically. An example would be a mono-shock system. Or a FV type system that has a spring/shock unit for bump and 2 shocks to control roll. Another example would be any 3rd spring system.

The number 1 right above that in the rules is you only get 4 dampers/Shock Absorbers. So sounds like you can do a Third spring system, as long as it's only a spring and no dampening. and yeah I suppose a Third spring system is kinda mechanically interconnecting left and right side .... but then as I just said you only have 4 dampers on the car, and it's the dampers that can't be interconnected so a heave spring alone without damping would be fine ... sounds like the rules make issue with it when there's a damper across that spring.

I'm more looking for specific confirmation I guess that the interconnection they're talking about is the dampers directly and specifically, that if they're "indirectly" interconnected through a linkage, (like my example above, one damper roll, one damper bump) that would be fine. Or what they mean by interconnected, it's not in the technical glossary.


That penske system your talking about. My professor when I did my Capstone project showed us a system schematic of something Yamaha was offering ... and It looked exactly like what you're describing

Huh, some guys like to run a shock absorber on their gas pedal. guess that only leaves them three for the wheels. I'm looking at Novembers rules I should update and see what language that arrived at (I assume it being written in red means it was just added to those rules). Yep there it is written in black now so was added in November. Maximum 4 dampers/shock absorbers. so you don't get to have one on your gas pedal anymore

Steve
I may be wrong about this but wasn't there a GCR rule change in 2015 that limited the number of shocks (dampers) per wheel to 1. I thought also that a third spring was no longer legal.

Might want to check that

Never mind I see the post above caught that in the GCR....

Quote:

---

Originally Posted by ghickman

Steve
I may be wrong about this but wasn't there a GCR rule change in 2015 that limited the number of shocks (dampers) per wheel to 1. I thought also that a third spring was no longer legal.

Might want to check that

Never mind I see the post above caught that in the GCR....

---

In a quick reading of the 2016 GCR, 9.1.1 Formula Category, G. Formula 1000, 9. Suspension, C. Shock absorbers. The only restrictions is what they are made from, steel or aluminum alloy body.

A further check it appears the FF/FC rules are essentially the same.

I don't think there are any restrictions on numbers or how the function.

If you want to have a linkage that connects the bellcranks, I am sure that you will be able to. I believe that most cars have an anti-roll bar of some kind.

Yes, there is a restriction in the general formula category against having more than 4 dampers. But, if you want to put one on the throttle pedal, you would almost certainly be legal. I believe that the intent of that rule is to control suspension development and cost.

Referring back to my previous post, the 4 damper rule does restrict you to 4 dampers maximum but I don't find in the glossary any definition for a damper.

Taken as a whole, 9.1.1.Formula Category does imply a definition. You will notice that this limitation only applies to "fluid dampers".

If someone wants to get anal, bump stops do work as springs and they are not made of steel.

Quote:

---

If you want to have a linkage that connects the bellcranks, I am sure that you will be able to. I believe that most cars have an anti-roll bar of some kind.

---

Quote:

---

I'm more looking for specific confirmation I guess that the interconnection they're talking about is the dampers directly and specifically, that if they're "indirectly" interconnected through a linkage, (like my example above, one damper roll, one damper bump) that would be fine. Or what they mean by interconnected, it's not in the technical glossary.

---

Quote:

---

You can use as many shocks as you want. The shocks can be interconnected mechanically. An example would be a mono-shock system. Or a FV type system that has a spring/shock unit for bump and 2 shocks to control roll. Another example would be any 3rd spring system.

---

Granted I'm cherry picking quotes here and the dialog has evolved however;


so we seem to agree that connecting a left and right bellcrank is fine say for anti roll or anti pitch or heave.

following that further, there used to be a suspension arrangement referred to as a Z-bar suspension. Where the FL and RR were connected to a torsion bar together, the FR and RL were connected to a separate torsion bar together. Nowadays running torsion bars down the center of the car would be a packaging nightmare, but we've all experienced a packaging nightmare with other mechanisms which are just connecting translating motion and the common consensus is hydraulics solves that problem very well. In my mind I'd consider a brake master cylinder and it's calipers still a mechanism, or a hydraulic shift ram a mechanism, or hydraulic clutch master cylinder and slave cylinder is a mechanism too, I think so at least.

If you had some horrible time trying to package an anti-roll bar could you solve the problem with the wonder that is hydraulics? a remote mounted anti-roll bar?

So at what point does a hydraulic connection for a Z-bar suspension arrangement become interconnected suspension and an Anti-Roll bar is not? If you'd allow the remote mounted ARB, and allow a Z-bar connection through Hydraulic remote mount means, then pretty quickly a guy could put the two together and end up with a Hydraulically Interconnected Suspension, Just 4 springs and dampers, but now they're each a suspension mode of operation rather than a wheel. The first year of competition you might use to the hydraulics to move a piston which compresses a physical coil spring over damper to explain what you've done and why you think it's legal, This is just taking the refinement of Gas accumulators out and replacing them with physical springs and dampers but you still get the same functionality. It's just taking what is already allowed to it's logical progression.

So is something Hydraulic no longer a mechanism? no remote mounted ARB's? no Z-bar arrangement without the torsion bars down the middle?


The wording singles out interconnecting across a shock absorber; What if you just had gas springs and no dampers? The line friction alone would give you some damping.


I know I'm reaching; Ever since you guys clarified what was meant with the "Shadow Rule" and how that seemed hidden in the language, It's definitely not the first thing you think of when you're reading that for the first time. It's left me questioning how I interpret what I'm reading more

It seems to me that you can get what you want by separating the springs from the dampers. You can connect the springs any way you want as long as your system does not have a damping function. You can go with torsion bars in place of coil springs or you could do something like the Shadowfax FF and use a collection of valve springs to support the car. Maybe have the chassis end of the springs interconnected with hydraulically movable mounts controlled how ever you wanted.

No reason that you can not have the springing system interconnected hydraulically But you can have each wheel movement damped with an individual shock.

Years ago one of my customers had an active suspension system that used engine oil pressure to activate a slave cylinder in place of a anti roll bar link.

I suspect that the rule is meant to prevent people running the front-to-rear systems that F1 was running a little while ago. If your system operates too closely to those, I suspect that you will run afoul of the rule.

You might want to call a CRB member and ask what they were trying to accomplish. There is also some kind of compliance verification process these days that you can use to make sure that something you want to do is legal.

But, if you really try to make something like that work, you are going to overcomplicate the car and you will never get done. Your time will be better spent just trying to get a car on track. Most people who go down the path you are discussing fail to even make it that far, much less to the point of getting their FRIC system working.

Believe an old Formula 1, 2, 3 spec ....... u r allowed to fair part of the tire, half the tire and only to the top of the rim, ........... why u see some pictures of the cars we love BACK when........ FSV was still in that realm although didn't necessarily exploit, FF had the trunk-width dealee for air down cheat control and other rules ....... believe that the Radon had too many whizzies and too much skin drag, balance that if u will ....... guy riding an engine, oh and wheels out there and that's it, no more drag, u get the drift......

Been missing your unique insight Modo. :thumbsup: :)

struggling Mike, never been 64 before, car is still in the trailer after blown engine last July, good thing is I can fix that, just fixed the tow vehicle, hobbled back from the SPoint in the Town & Country, ..........maybe make the school at NJ in August to renew per requirements, could still be licensed year to year since 91 and 5 years in the seventies ....... it's silly, I like still being a participant in an organized sport in my 60's, yea, auto racing a little different but still .......... hope to see you out there in the near future .....

watching Indy at St. Pete now .......... boy does that and the warmer weather make youi want head to the garage, LOL ......... right past the laundry room door to get that Italian chassis back out there!!!!

I just accepted employment making cars on someone else's dime, making his car ideas. Soooo don't think I'll have time to carry on with this. Shouldn't be a problem enjoying these new cars with Dad so it's all working out; and who knows this new employer hasn't completely ruled out an F1000.

Thank you everyone for your insights. Everything has applications in car building in general and I carry these lessons with me forward. Again, thank you everyone.

AlbertaSpeedShop