Suspension pickup positions belong to the suspension, and the frame's job is to get to them.
Get that order backwards — draw a nice chassis, then hang suspension off it where there's convenient structure — and you'll end up with geometry that handles the way the structure happened to be shaped. Which is to say, badly, in ways that are expensive to correct once there's steel.
This tutorial covers establishing those points as fixed constraints before any chassis member exists.
What a hard point is, precisely
A hard point is the location of a pivot axis, expressed in your project coordinates, that the suspension design requires and the chassis is obliged to deliver.
For a double wishbone corner that's four of them per side: upper front, upper rear, lower front, lower rear. Plus the steering rack inner pivots, the damper or pushrod mount, and the anti-roll bar mounts if you have one.
The point is the pivot axis, not the bracket. The bracket is a fabrication problem you solve afterwards. If you start modelling brackets now you'll anchor the geometry to whatever you can easily make, which is the mistake this whole tutorial exists to prevent.
Where the numbers come from
Three sources, in descending order of how much you should trust them.
A suspension geometry package. If you've designed the kinematics properly, the pickup coordinates come straight out. Transform them into your project coordinate system and you're done. This is the right way.
Measured off a donor. If you're adapting an existing subframe or arms, measure the real thing. Measure to the pivot centres, not to the visible bracket faces, and measure at a known ride height with the suspension at a known position. Write down what that position was, because a measurement taken at full droop and used as if it were ride height is a very confusing error to chase later.
Scaled from a reference model. Workable, and the weakest option. Verify the scale against two independent known dimensions first, because an error here propagates into every corner. The measuring tutorial covers the procedure.
Getting them into the model
Place a joint at every hard point before you draw a single member.
You'll have a cloud of unconnected joints floating in space with no structure between them. That looks wrong and it's correct. Those points are the specification; everything else is the solution.
While they're still isolated and easy to work with, do both of these.
Name them systematically. "LF-UPR-FWD" beats "Joint 17" by a margin that grows with every hour you spend in the project. You will be reading these names on a fabrication sheet at some point.
Apply symmetry immediately. Suspension points are symmetric about the vehicle centerline, and the moment you mirror them you've halved the work and guaranteed the two sides can't drift. Do it now rather than after you've built structure around them — retrofitting symmetry to an already-asymmetric set of pickups means deciding which side was right, and you won't know.
Check them against each other before building
Isolated points are easy to verify and hard to verify later. So do it now:
- Measure the distance between upper and lower pivots on each corner. Compare against your intended upright. If they don't match, nothing downstream can save it.
- Check left against right. With symmetry on this should be exact. If it isn't, symmetry isn't actually applied.
- Check the pivot axis angles. The line through the front and rear pivots of a wishbone defines its axis, and that axis determines anti-dive and anti-squat. It's rarely parallel to any vehicle axis, and it's easy to get subtly wrong.
- Check front to rear. Wheelbase, measured between the points you actually placed rather than the number you intended.
Ten minutes here. Each of these is a problem you can fix by typing a number now, and a problem you fix with a cut-off wheel later.
Then build the frame to reach them
Now add structure, and the rule is that the frame comes to the points. The points do not move to suit the frame.
Load paths at a pickup matter more than almost anywhere else on the car, because the loads are high and they're cyclic. Three things to hold to:
Land pickups at nodes. A suspension load arriving mid-span on an unsupported tube bends that tube, every cycle, forever. Split a member to create a real joint at the pickup rather than hanging a bracket off a convenient face.
Triangulate the pickup into the structure. The load has to reach somewhere stiff. A pickup on a well-braced node is a different component from the same pickup on a node that's only connected in one plane.
Keep centerlines intersecting. Where members meet at a pickup node, offset centerlines introduce a bending moment exactly where you least want one.
Then check it moving
Static clearance at ride height means nothing. Articulate the suspension through full bump, full droop and full steering lock, and check for contact at every position — not just the extremes, because the tightest point is frequently somewhere in the middle of the travel where an arm sweeps past a diagonal.
Then run the simulation with a representative corner load and look at how the pickup node behaves. Deflection at a suspension pickup is compliance, and compliance there shows up as geometry that changes under load. The car will understeer on turn-in and nobody will be able to tell you why.
If the node moves more than you're comfortable with, fix it with triangulation before you reach for bigger tube.
Comments
Loading comments…