I think the image of the watts linkage in conjunction with a De-Dion system below may help you.
I'm midway through converting my car from live axle, to de-dion. As there is no metalwork beyond the axle centre-line, just fibre-glass bodywork,
I'm making a subframe to mount the trailing arms and diff. I also decided to make a Watts linkage for sideways location. As the diff is now
mounted to the subframe, it made sense to mount the centre of the Watts linkage to the back of the diff with the arms attaching to the de-dion, one
above, and one below.
Normally, it's the axle that carries the centre of the Watts linkage with the arms attaching to the chassis.
But something was niggling my fuzzy little brain. Something about the way the centreline of the axle would move in relation to whereabouts on the
suspension stroke the axle was.
Enter my sons Lego......
Sure enough, the Watts linkage holds the cen
treline in relation to the points where the outer ends of the arms attach. With them attached to the axle, the axle moves right on full compression
and left on full extension.
At least I found out before I got it all welded-up!
Meccano, now that's the stuff.
quote:
Originally posted by matt_gsxr
Meccano, now that's the stuff.
Hi
I think the image of the watts linkage in conjunction with a De-Dion system below may help you.
Cheers Hector, that piccy is of the traditional way of fitting a Watts linkage. My idea was to pivot the link behind the diff, in front of the axle, mounting the arms to the axle instead of the chassis. The Lego tells me why my idea won't work!
It doesn't matter whether the watts linkage is mounted as your proposing or the more conventional way, as the motion of the watts linkage is not
truely a straight line, at extremes of travel both versions of the watts linkage will deviate from the centreline
Cheers
David
[Edited on 25/2/12 by Minicooper]
If the movement of two components relative to each other are parallel, then which ever component is held static the other component will still move
parallel.
I went through the same logic process as you when designing my dedion suspension. I looked at the 'reverse' mounting option too. I concluded
that the watts linkage was not practical in reverse but it would still work.
Are you sure that you set the two arms level when suspension is at rest?. They must not go to the centre of the tube but above and below at the same
vertical spacing as the centre swivel.
Hi
You also have to make sure that the watts linkage pickup points on the chassis / sub-frame are not flexing in relationship too the main chassis other
wise you have as much sideways movement as a pan hard rod setup. Have you got weights from before to compare to after on the conversion.
Cheers Matt
It does matter which way round the Watts linkage is mounted, that was the point of the Lego! I'll try to get it off my son again and video what
happens. I have the arms parrallel and horizontal at rest, one above/left, the other below/right.
The Lego will show what I mean.......
As Minicooper says, the travel is not going to be perfectly centreline whichever way it is set-up.
The vertical displacement of the centre swivel is the same as the vertical displacement of the arm mounting points. This means that the two arms
follow the same arc initially and so foreshorten by the same amount and so the travel of the swivel is initially on the centreline. Gradually the
centre swivel's rotation means that it's vertical displacement reduces and one arm starts to move through a greater arc than the other,
moving the swivel and therefore the axle off the centreline.
This will happen whichever component is fixed. The motion at the extremities can be improved by slightly reducing the displacement of the arm mounting
points,thereby inducing a slight 'S' shaped travel. The watts is still vastly superior to the panhard anyway.
I am using a 6 link set-up with two diagonal 'Satchell' links to give lateral location.
quote:
Originally posted by plentywahalla
Are you sure that you set the two arms level when suspension is at rest?. They must not go to the centre of the tube but above and below at the same vertical spacing as the centre swivel.
Here's a vid clip to try and show what I mean. As the centre of the Watts stays parrallel to the ends of the arms, when the axle tilts, ie going
over a bump or cornering, the centreline moves in relation to the body. Sorry for the jibbering in the clip.
https://www.youtube.com/watch?v=IDb7t75a4iI&feature=youtube_gdata_player
However, as there will be hardly any roll in the cars rear due to the poly bushes in the trailing aems and coilovers, the off-centre thing will be
negligable. So I'll mount it which ever way is easiest!
I'm still not convinced despite the video.
I have done a quick cad plot of the trace followed by the centre swivel which shows it tracking right during compression and left on extension. It
follows a very gentle 'S' shape with the centre portion virtually vertical.
If you compress just the left side i.e. if you hit a bump, the it exaggerates the sideways movement as the centre portion of the 'S' trace
is tilted to the right. If you compress the right hand side the 'S' tilts to the left which effectively negates the sideways movement as the
initial travel is now leftwards and on further movement it tracks back to the centreline.
If I understand your Lego model what you are demonstrating is the difference between the above conditions. That is how the axle reacts to opposing
bumps or body roll. One way it reacts better, the other way is worse. This condition can be improved by reducing the displacement of the axle mounts
to about 10% less than the length of the swivel.
If you mount the watts linkage back to front, it doesn't change the reaction to roll or bump, it just reverses it.
I didn't go with the 'back to front' diff mounted swivel because it increases the distance between the tube and the axle centreline.
This distance wants to be kept to a minimum to reduce the torque reaction on the tube.
[Edited on 26/2/12 by plentywahalla]
The reason it looks so bad when reversed is that any Watt's link only has a certain amount of straight line travel, if any, and you're
pushing the axle past that limit. You've got it set so that the bounce is more than the wheel height. Any road car will already broken if you
managed that anyway!
As long as the links are parallel, then the movement must be identical which ever way round they move as they only move in relation to each other and
as such can only move in one way. The only way to lengthen the straight section of vertical movement is to increase the swivel length, ie the longer
the swivel, the longer the straight trace will be.
This shows the difference:
Cheers for the input guys.
The subframe/diff/axle/Watts is all tacked together now so I've been able to bench-mount it and move to through it's full range of travel.
The only sideways movement with the axle level, is the few mm of movement in the nylon bushes. With the axle at a few degrees from horizontal, it
moves 6mm out of true with the subframe. Not enough to worry about!
I made a slight error with the distance between the front of the axle and the back of the diff but if it looks ok once on the car, I'll make a
slimmer Watts pivot link.
on my axle (designed by by SHP engineering) has the arm attach to the axle and the pivot on the chassis. and i'm happy with it.
yes the axle will move in a slight S shape, moving side to side on full bump and compression, but the movement is going to be the same either way. the
centre pivot will stay centred in relation to the arms as long as its set up properly. so, if the axle remains in the centre, the chassis will move in
the same pattern regardless of which part the pivot is on.
In your lego demo (videos are very jumpy on my laptop for some reason) were both arms set up completely parallel with each other, or were they angled?
were they both the same length?
heres the important bit
yes it shows the pivot moving up and down, but it works equally the same if the arms move up and down and the pivot stays in the centre (but of
imagination needed on seeing the picture move that way)
if yours is pivoted on the axle and the arms attach to the chassis and it works, then perfect, but either way will work with the same outcome tbh
here's mine
i messed up a bit on measuring the bars out, but they are equal length and thats whats important, i'll redo them later anyway, and the axle
isn't in the centre in that pic as i just took the photo as soon as it was put together
[Edited on 27/2/12 by blakep82]