This article was first published in Autospeed.
The thing that really gets me excited and interested in constructing Human
Powered Vehicles is that I think they are immensely hard to design. But,
paradoxically, after the design is done, you can make one in your backyard shed!
The main engineering problem is that they need to use an approach which is low
in weight. And that makes things really, really hard – especially if you want to
achieve good results in other aspects, like ride and handling!
Take for example the front suspension. In my first Human Powered Vehicle I
used an unequal length, double wishbone front suspension design. (The arrow
points to the lower wishbone.) That sounds really good, but even with shorter
upper than lower arms, a double wishbone design will weigh something like 50 per
cent more than some other suspension designs. Why? Well, it’s got two lots of
Hmm, OK then, so maybe this isn’t quite as simple as it first sounds. Let’s
take a look at the criteria and see what alternatives are available.
Why is low weight so important? More than anything else (yes, even more
important than getting unsprung weight down) it’s because suspension contributes
an (un)fair proportion to the total weight of the machine. And when you have
only one human being to power it, heavy suspension makes the power/weight ratio
worse and worse...
Wheel travel of at least 100mm
As we’ve covered in previous parts in this series, and as I found with my
first design, a road-going HPV needs at least 100mm of suspension travel. More
would be nice, but I don’t think less is a viable option.
Dynamic Camber gain in bump
Dynamic negative camber gain is needed so that when the machine rolls in
cornering, the wheel stays closer to upright. Furthermore, the thrust developed
by the camber angle aids cornering. If the HPV is to roll (say) 5 degrees in
full cornering, something of the order of 5 degrees per inch of suspension
travel is typically needed.
Low Motion Ratio
A low motion ratio means that there’s not a lot of leverage over the spring –
ie, that the wheel travel and the spring travel are similar. The reasons for
this have been covered in previous parts of this series: suffice to say here
that it reduces frame input loads and makes the selection of a spring (and using
that spring) easier. In fact, I am using Firestone industrial airbags as the
Now you might have skimmed all of those criteria, saying to yourself, yep,
But what’s the answer? What suspension design would you pick? That’s a helluva lot harder...
Let’s take a look at the different types of suspension that could be used,
judging them against the above criteria. The basis for the following discussion
is mainly drawn from
Front Suspension Designs
, an article I actually prepared when thinking through my first HPV
Solid front axle – this design is poor because it doesn’t have
dynamic increase in camber and allows the up/down and steering
motion of one wheel to affect the other. Positives are that it’s a simple
suspension that can be very light. Huh? But aren’t solid axles heavy? They don’t
have to be, especially if the springs are placed as far outboard as possible and
so the bending loads are reduced. Significantly, a solid axle allows a much
simpler steering system to be used (ie bump steer more easily avoided) than an
independent suspension design.
Sliding Pillar (or Sliding Kingpin) – it’s difficult to get 100mm of
suspension travel (although not impossible) and there’s no dynamic camber
increase. The frame support that takes the spring load has to extend right out
to above the wheel (and so is heavier). However, since most HPVs run kingpins,
it’s easy to integrate this suspension design just by using longer kingpins and
steel or plastic springs between the kingpin and frame support. There are also
no additional suspension arms, making it simpler and lighter than most
alternatives. However, avoiding bump-steer with sliding pillar suspensions having a lot of travel is near impossible.
Leading Link – with this approach it’s hard to get long suspension
travel and the design uses lots of pivot points. Again, the frame support that
takes spring load has to extend right out to above the wheel (and so is
heavier). However, bending loads in the suspension arms are at a minimum (most
are in compression) so they can be made lighter.
Swing Axle – the high roll centre of this design creates jacking forces,
and with 100mm of wheel travel, the dynamic camber gain is likely to be too
great without very long suspension arms (and so a wide track ... which would
make for a heavy HPV).
Trailing Link – it’s hard to get 100mm of wheel travel with short arms
(and long arms are heavy – and there are two of them, remember!), there’s no
dynamic camber increase in bump, and the design is best suited to torsion bar
springs (which have a high motion ratio, so making them heavy).
McPherson Strut – despite being used in so many cars, this design has
poor dynamic camber gain, and frame support that takes spring load has to extend
right out to above the wheel (and so is heavier). However, as with Sliding
Kingpins, it can be integrated more easily into existing suspension designs and
the suspension arms don’t need to take up much room within the track.
Wishbones – As alluded to above, because there are two of them (upper and
lower), it’s hard to get the weight down. And it’s not just the weight of the
arms themselves – normally there are six pivot points, and even if using
lightweight rose joints (rod-ends or Heim joints – they’re all the same thing),
the weight certainly adds up.
Where to From Here?
Now if you hadn’t realised, that’s every suspension system
covered. (Of course there are other variations on these fundamental
designs but they add complexity and weight – eg they use additional suspension
arms in a multi-link approach.) Of those described above, the ‘winners’ are not
at all what you’d first think if you’re familiar with car technology.
In fact I’d rate the top three as:
1. Solid front axle
2. Swing axle
3. Sliding Pillar (or Sliding Kingpin)
However, because of the deficiencies cited above, none of these approaches
excite me. Trying to come up with workable alternatives I wracked my brains for
hours; I grew grey hairs that then fell out; I read numerous books (mostly 60 or
70 years old: they are by far the best on fundamentals of vehicle design) and I
struggled and struggled. What could I do that would give increasing negative
camber in bump, be lightweight, have at least 100mm of travel and use an
Next week: what I did.
To show how much I’ve learned, you might want to take a look at
Building a Human-Powered Vehicle, Part 1, and especially the
break-out box at the end. That’s right, some of the very same designs I
rubbished there are amongst those I now think potentially the best! (Although, I
must still say that I think the execution of the designs described in that
article are poor.)
Did you enjoy this article?
Please consider supporting AutoSpeed with a small contribution. More Info...