The Mark Ortiz Automotive
Reproduction for sale subject to restrictions. Please inquire for details.
Mark Ortiz Automotive is a chassis consulting service primarily serving oval track and road racers. This newsletter is a free service intended to benefit racers and enthusiasts by offering useful insights into chassis engineering and answers to questions. Readers may mail questions to: 155 Wankel Dr., Kannapolis, NC 28083-8200; submit questions by phone at 704-933-8876; or submit questions by e-mail to: email@example.com. Readers are invited to subscribe to this newsletter by e-mail. Just e-mail me and request to be added to the list.
SIGN CONVENTIONS FOR CAMBER GAIN AND RELATED QUANTITIES
Last month I was discussing camber gain, in response to a question. I said that camber gain is generally considered positive when camber goes toward negative as the suspension compresses. I also introduced the concept of camber acceleration, which is the rate of change of camber gain with respect to suspension displacement, and I said this would be positive if camber gain increases as the suspension compresses.
It should be understood that the term “camber gain” originated with racers taking measurements on a car in the shop, not engineers creating math models. To make the sign conventions really make mathematical sense requires some thought.
Far as I know, there are no official SAE or ISO sign conventions for camber gain or camber acceleration, but there are for suspension displacement (or “jounce”) and for camber. Per ISO convention, suspension displacement is positive in compression and camber is negative when the top of the tire is inboard of the bottom. To be consistent with these conventions, camber gain as we normally use the term would properly be called absolute camber gain, and camber gain as a signed quantity would be negative for most suspension systems.
Correspondingly, camber acceleration would be negative for most SLA suspensions. Camber gain would decrease (i.e. increase negative) with respect to displacement (compression being positive). Camber acceleration would be positive for a MacPherson strut.
A LITTLE MORE ON TRIKES
Last month, when discussing trikes, I mentioned the Trihawk as an example of the right way to do a three-wheeler and said whoever designed knew what they were doing and got it right. Turns out credit should go to Bob McKee for this. I’ve known about McKee since the ‘60’s, but either forgot
or never knew about his involvement with the Trihawk. Thanks and a hat tip to Doug Milliken for filling me in on that.
Bob McKee is still around, and appears in a recent “Petrolicious” YouTube video (https://www.youtube.com/watch?v=GAgLSy371us) about one of his better known designs, the Howmet turbine car.
Also, Peter Olsson had some information on leaning sidecars:
“The passenger is indeed controlling the lean of the motorcycle. Note that the sidecar is on the right side of the motorcycle, not the left as is the norm. These racing outfits were 500c.c., usually British single cylinder engines. The rider had no control over the lean; the right foot is being used purely for bracing. These machines would all have had right hand gear-changes; once off the start line, there was no requirement for gear-changes, so the right foot could be used for anything else. Whether these "leaning sidecars" were outlawed, or simply fell out of fashion, I do not know. The photo appears to have been taken prewar.”
DESIGN AND ANALYSIS METHODOLOGY
I have been reviewing several of your past newsletters on live axle suspension design, in particular, the one from September 2007 on the design of the Satchell Link Rear Suspension with comments from Mr. Terry Satchell.
It occurs to me in reading many of your writings coupled with Mr. Satchell's comments in your newsletter mentioned above, as well as his chapter on Suspension Geometry in RCVD that what is commonly referred to as the 2 x 2D graphical approach can be used very effectively in designing a 3D suspension linkages. I have read elsewhere many comments that disparage this approach with the naysayers claiming this is a 3D problem and should only be solved using either 3D graphic or 3D analytical methods.
I use Solidworks 3D modeling in my work as a tool and model maker daily. But as an amateur race car engineer software packages such as this are simply out of reach due to their high cost as well as any really good kinematics software.
I was trained at the drawing board, have a good knowledge of descriptive geometry and have come to believe that my 2 x 2D drawing board or 2D CAD solutions are accurate enough for most purposes.
What is your opinion of the 2 x 2D graphical approach to designing and analyzing suspension kinematic motion?
In plotting solid rear axle displacement in heave or roll when using a Panhard bar how would you handle axle scrub? This seems to be one area that must be iterated to be exact. However the longitudinal link movements caused by scrub seem, in 2D CAD, to be rather small.
What parameters would you track other than the usual movement of: right and left instant centers, roll center, axle roll axis and wheel path?
Realizing that the axle roll axis is an engineering approximation and becomes more so as the suspension moves, would you consider your approach of averaging the divergence of link intersection points in side view adequate over the full bump/droop range of rear axle travel to fix points to construct the axle roll axis?
I am not adverse to, or uncomfortable with, using analytical methods. I just find a graphic approach seems to get at a suspension design quicker and with a better 'feel' for how it works.
Does Terry Satchell work primarily graphically in laying out and then analyzing his suspension designs?
Do you work more graphically?
Is it simply a romantic notion that Adrian Newey still works at the board even though we know he is backed up by a small army of engineers and designers using the most state of the art computing and software. Or does it just make a good image story?
I don’t know Adrian Newey, although of course I know who he is and have likewise read that he prefers to design manually. I did put the question to Terry Satchell, although I know that he’s officially retired. This is what he had to say:
“I would be glad to answer this guy’s question. And to confirm, I am fully retired. I go to the gym in the morning and ride my horse in the afternoon. I am enjoying that schedule.
“I would have to say that what I do is analytical rather than graphic. I have developed some Excel spreadsheets that quickly solve for suspension parameters from pivot points I put in. And I also use the iterative solver in Excel to "synthesize" the geometry points from the desired static design factors at layout position. Once I have that the way I want it, I then transfer to a bespoke 3D geometry program that gives me the factors through suspension travel up and down. I then fine tune from there by modifying the appropriate pivot in the appropriate direction to get what I want. This is for independent suspensions.
“For live axles I have another bespoke 4-bar link analysis program. To get started with this type of suspension, I do some simple hand calculations to get the starting points then iterate the 3D geometry program.
“In all these cases I am assuming I have knowledge of some of the pivots, for example the knuckle in an independent and the axle end for a live axle rear. I get them from drawings, or previous usage of the same parts, or we would CMM the particular part. When needed, I have taken CMM points of a component, and had a CAD ‘friend’ help me put the component in ‘car position’ at layout design. Then I use the process I outlined above.”
In my own consulting work, I am seldom called upon to design a suspension, or to model or simulate an existing one. I am mostly asked to answer specific questions, solve specific problems, or make specific recommendations. Sometimes these are for cars in the design stage, but more often they are for cars that already exist. The inputs I get vary a great deal, and so do the outputs the client needs from me. In the vast majority of cases, the car is either a purchased race car or one that the client race prepared or built without drawing or CAD modeling it first.
Sometimes I get called in by engineers for major teams, but mostly I serve hobbyists and lower-level professionals who don’t have in-house engineering staffs. Accordingly, they generally don’t have drawings or CAD models of their cars, nor do they have surface plates, height gauges, or CMM’s. They are generally at a remote location and are working with me by phone and e-mail. Sometimes they have purchased geometry programs. Sometimes they can tell me where their instant centers are or what their camber gain is. Most times I have no way of verifying their numbers. It is not uncommon for their measurements or observations to contradict each other.
In many cases, I can get what’s needed from verbal descriptions and photographs.
I call what I do brain aided engineering. I understand suspension in depth and consequently am able to ask the right questions to address a client’s specific question or need without having to know everything about the car or comprehensively model it. I am not constrained by a set method or software package.
I do sometimes work with students who are doing designs. They generally use Solidworks. For SLA suspension, I generally have them start with static force lines and instant centers. If we are trying to use existing uprights, we use the existing ball joint locations and the desired instant centers to establish control arm planes, and then see what we can come up with that packages with the rest of the car.
When I do design myself, I work manually. I have a 72” x 42” drafting table with a Vemco V-track drafting machine. I’ve owned it since 1982. There are things it can’t do, but it’s 100% reliable. It never crashes. It has no compatibility problems. It’s paid for. There are no licensing fees. I don’t need to hire anybody to help me. I can use it not only for drawings and geometry studies, but also for small part layout and making cardboard templates. If mistakes are made, they’re my mistakes, and I can find them and fix them. I can live with that kind of world.
Where computers are needed the most is suspensions with multiple links where no links are in a common plane and nothing looks like a control arm. If you are intent on designing something like that, you pretty much have to do it iteratively with a computer. Once you have an iteration to analyze, you can find instant centers and force lines analytically.
However, be aware that computer analysis is only as good as whoever wrote the program. If you see a commercially available program whose originator makes serious errors when conversationally discussing the systems at issue, you can be pretty sure their simulations will be inaccurate. I can’t help noticing that despite the availability of ready-made simulation programs, I am constantly seeing ads from big-name teams for engineers to write proprietary simulation programs just for them, which they then guard closely as trade secrets. Nature is trying to tell us something here.
You need to be able to sanity check the outputs. For example, one time I was working with a widely used and well respected geometry program and looking at its outputs for bump steer. This was for an SLA front end with a front mounted steering rack. When I moved the rack up, the program told me I was increasing toe-in with suspension compression. Basic mechanical reasoning tells us that’s impossible. It’s important to be able to do basic mechanical reasoning and not just blindly trust the computer. The computer may give you answers to five decimal places, but if you can’t check the outputs with your own intelligence, the truth is that you really don’t know if the computer output is correct or not.