I have been investigating the stiffness of the MX-5 chassis, especially the 2006 Mk3 which I own. From sources on the web, I made the comparison in the following chart. I have picked all the data on MX-5s and cars at the flexible end and the stiff end of the range for convertibles and roofed types. I then accounted for the need for heavy and wide cars to be stiffer. The notorious Ford Mustang convertible is effectively half as stiff as the nearest rival,; how awful must that be to drive around a corner? At the other end, the RX-8 is amazing, some folks on this forum must have experience. The Lotus Elan has a low absolute stiffness but effectively much stiffer due to low weight. And the Saabs clearly show the effect of cutting off a roof. As for MX-5s? Well, they start out down with the Nissan Micra, but later NBs improve on this, and NC and NDs are in the best convertible class, with the ND being the second most rigid convertible in the full data set. Mazda certainly learnt a lot over 20 years.
Yes, of course, there are many caveats to apply to this but it is mainly for interest. Of course, many of you are thinking “why did he waste the time doing this when he could have been driving the car”.
interesting chart!
I found that my NC tin-top was way too flexible! so I added chassis braces and it’s now alot better.
tbh, it’s transformed the car.
in my opinion, braces are an absolute must-have mod for the NC
This is an interesting graph; but I am not sure how to interpret it in absolute terms. Where does one draw the line between safe and unsafe? I am quite prepared to imagine that the Mustang is a death-trap; but Is the ND good enough to be on the acceptably safe side of the line?
David
As cars have suspension that moves I’m not convinced you can draw much of a conclusion from just the stiffness of the car. The Tesla is pretty stiff but unlikely to handle better than a less stiff Aston Martin.
In my opinion, the safety depends how you drive, so there is no line. The “soft” end requires driving with slow changes in direction and avoidance of cornering at speed or over rough surfaces.
I think the biggest difference is that suspension has damping and hopefully well controlled changes in wheel geometry when cornering and travelling over uneven surfaces. Probably, the twisting of the chassis is relatively undamped, i.e. there could be several oscillations after the initial bump or turn in, and the wheels can have lower contact when the chassis twists.
I have driven cars in the US where the suspension is also pretty soft. I once had a GM equivalent of a Vauxhall Cavalier, and thought I knew how to drive it, until I tried to make a late turn onto an off ramp, resulting in an awful oscillation, and a missed turn.
That’s interesting to hear; which brace did you install?
nearly all of them!
the NC comes from the factory with the big brace underneath just behind the engine bay. but thats all they have.
i added the brace that goes across the top of the rear shocks that is in the boot!
(it stopped most of the roof creaking) i also added all 3 rear bottom chassis braces.
all of them were the IL Motorsport ones sold by mx5 parts
they have made a huge difference to the car, both in the way it feels and drives, they have also stopped all of the roofs creaking and the roo9f also closes straight even if the car is on uneven ground
if you check my project thread and scroll down to the post from the 3rd of December. you can see the boot brace in those pics and then keep going and you will find the pics of the underside braces
post dated March shows 2 of the braces and then the second to last post shows the x brace
On my old NA, you could see and feel the dash flex when the LH front wheel.hit a hole. That doesn’t happen with the NC.
Its a bit selective. The 2013 Mustang was nothing more a refreshed early 80s F-Body. The 2016 model was completely new. The source is dubious; it seems to be US front for a Chinese analytics company. Basically they will scrape the internet for data, but not really understand the data. Hence there are two Elans featured, but neither will be the original Elan. I suspect they have pulled two sources giving different values for the M200 Elan.
There are few sources given. Some sources are manufacturer press releases claiming “X% stiffer than the old model”, which will be a quote designed to be the most flattering.
The “MX-5 NA FL”; this is the Mk1 with the thin bar behind the seats, which somehow ends up being stiffer than the NB, with its cockpit gussets and foam filled sections.
I think the underlying data is bad.
I agree that the data should be treated with a lot of skepticism. Especially as there is no indication of where the moments and rotations are applied and measured, so they may not comparable. The MX-5 NC data I feel most confident in because it matches the technical review from Mazda.
I have taken measurements on my 2006 NC (sans FSTB) and I find only 1/3 of the stiffness given in the Mazda document, when measuring at the inboard ends of the lower suspension struts/wishbones. The chassis twists by 0.6 degrees when applying 2000Nm, that is applying 300kg force at the suspension mount.
This is not explained by the missing brace, as the tower tops only move by c. 1mm when applying the same moment (equivalent to 1/10 degree). It is unlikely that a brace would actually reduce much of the motion at suspension mountings because locally it only effects the extension of the springs, but globally it may increase the torsional rigidity of the whole chassis structure.
So I plan to design and make some braces myself, and see what happens.
Somewhere else on this site I got into a conversation over strut braces. My ND is fine with the roof up but once down it does judder. I can also feel it on the steering wheel. I know a lot of people say it’s chassis twisting. I understand that but that twist must continue up the scuttle & the steering shaft goes through the scuttle, firewall whatever. Not sure if there’s a bearing at that point. I had this on a Vauxhall Astra when hood was down. Why does Mazda fit tower brace back to the scuttle on some models. II know some will disagree with me but I’ve fitted a DaveFab strut today not tested it yet though. Really light & welding is perfect. I had to replace the panel on the scuttle & also the brackets on the towers but not a bad job took about 3 hours.
It’s been said before, but:
Park your MX-5 with the roof down in a wide space.
Open both doors.
Stand well back, crouch down, and look at it from the side.
Be amazed at how little there is holding the front and back wheels together.
Convertibles are a real challenge.
Read somewhere that the gusset on the NB’s tunnel-to-rear bulkhead improved torsional rigidity by 40%. My NB improved, for a while, with new door bushes. Marked difference, but about 1000m later back to lettuce leaf.
The roof must lock in really tightly to achieve this, because the twist motion is less than a mm. My tests suggest that the OEM fully removable hard top for the NC does not provide any additional twist resistance.
Yes, that is a great image! But the NC has 5 separate torque tubes in that narrow floor, especially the transmission tunnel, which is why there are braces across the tunnel and back and front to close the tube. On mine, the rubber blocks in the doors do not fit tightly enough to increase to increase stiffness, but I am going to work on that.
Yes, the tunnel is the key to the stiffness, as per the Lotus Elan “back bone”.
As a driver that has two two seat sports cars which are close to eighty years apart in design
one with chassis that works as part of the suspension the other a modified Eunos roadster you have to drive both differently the 1933 car is low powered and can be made to corner suprisingly quickly on it 3.5 inch wide tyres but bumps unsettle it.
The Eunos will corner faster than is sensible on the roads. Both cars are used on the road and in sprints and hillclimbs.
Most drivers will never drive a car hard enough for a bit of flexibility to be a noticable issue.