The Mirage Clutch (analysis & weaknesses etc., with pics)

[Loren]

Okay, I'm no clutch expert. I've replaced more than a few of them, but usually because they were flat worn out or a clutch related part failed (pilot or throwout bearing, oil seal, etc)... or maybe for a performance upgrade, or more often a "just because the engine is out" replacement. And the important part, I've never really looked at them all that closely. Sometimes they're worn, sometimes they're not. Never really gave it much thought, but mostly just looked at friction surfaces. I guess I'm just getting more curious in my old age.

I did learn a couple years ago with a lightweight Miata project that an "unsprung hub" clutch disk is a pretty awful thing to drive. Those little springs in your clutch really do have a purpose! We'll come back to that.

The purpose of this thread isn't really to provide any conclusions. Merely some discussion, food for thought, and perhaps enlightenment based on my measurements and observations of the Mirage clutch that I happen to have on my kitchen table. (and the wife hasn't complained... maybe she finally gave up?) Opinions, experiences and feedback are welcome.

So, what I have before me is a 40,000 mile Mirage clutch. The first 30k miles happened over 2 years, likely a highway commuter vehicle. The last 10k were all me over the past 15 months, mostly casual city driving, and 35 autocrosses (214 autocross runs with appropriately aggressive launches and hard 2nd gear shifts).

PRESSURE PLATE:

I've already sent the pressure plate out for recycling, so no photos. I'm replacing it just because it's out of the car. For $105, it's always worth replacing a clutch while it's out to extend the life of the clutch. The pressure plate was nearly perfect. The spring fingers had significant (but normal) wear where the throwout bearing makes contact. That surprised me a little bit, but they would go to at least 120-150k miles at their current wear rate. The friction surface was barely broken in. The factory milling marks were still very present in the middle of the contact area, only the outer edges were really worn-in. Never any problems with the clutch slipping, which means the clutch itself is probably slightly more than needed for the amount of power we have. (really surprising for ANYTHING to be that over-built on this car, the tendency is for everything to be as light as possible and just barely adequate for the extremes of what the car is intended to do)

FLYWHEEL:

Uh, well, honestly... I didn't look at it too closely. It definitely wasn't showing enough wear that I even gave it any consideration. Probably about the same as the pressure plate. It's been sent off to get lightened.

FRICTION DISK:

Ah, now this is the fun part. This car has always had a bit of a noisy clutch. Just a bit of a rattle at idle, and a horrible resonant buzz if you get on it too hard in reverse. (used to do it every time I backed up my driveway until I figured out that it didn't do it if I babied it) I figured I'd just not worry about it until it became a problem, got a lot louder, or I pulled it out to put in a lighter flywheel. May hand was forced when an engine problem required the engine to be removed... time to do the flywheel and clutch!

The only clutch problems I've heard of from this generation of Mirage have been noises similar to mine, and one clutch with a broken spring in the clutch disk. So, the first thing we're looking at is the clutch springs. There are four springs. Actually, eight. Each of the main springs has a smaller spring inside it. All of the springs are quite loose. They rattle, they rotate easily, and they can be moved foreward and back about 1.5-2mm. There's the noise I've been hearing.

But, that's not the worst part. Rotating the springs reveals that the springs are quite worn. Now, I don't know how much wear is "normal", but there's enough wear on these that I would expect at least one of them to break by 80k miles. The deepest grooves measure about 1.2mm. (the spring material is only 3mm thick)

BASIC SUMMARY:

Pressure plate and flywheel are fine. Very little wear for 40k miles. The clutch disk friction surface is also fine. The springs in the clutch disk are possibly a problem area, especially for an aggressive driver.

I'm gonna add two more posts here. One to expand on how the clutch disk works and what I think happened or is happening. (photos included) The other to compare the replacement clutch disk with the old and provide some interesting measurements that you're never going to need.

[Loren]

What's up with the cluch disk?



The stock clutch disk is made by Exedy. They make clutches for a LOT of new cars, and are an extremely common aftermarket replacement option. Definitely a "standard" among clutch manufacturers, so that's good. But, this is definitely not a "performance" clutch. It has at least one area of weakness that is readily apparent to me after a year of autocross competition. Mind you, a year of autocross for me is equivalent to about 3 years of what a typical "casual" autocrosser would do. I do at least two events per month, 12 months per year, often with other people taking runs in the car, etc. So, if I've done 15 months of autocross... and it seems I'm saying that the clutch would have failed in 30 months... that's with MY use. A typical autocrosser would get 3x as much out of it... that's 90 months! So, please don't think I'm saying the clutch is junk. It's not. And don't be afraid to autocross your car (at least don't be afraid for THIS reason... the car IS prone to rolling over when pressed due to it's top-heaviness)

Parts of the clutch disk...

FRICTION SURFACE - On each side is the "organic" friction surface. I don't even know what it's made of. It's like fiber-reinforced resin of some sort. Designed to take a lot of abuse with minimal wear. It's riveted to the marcel, and plenty adequate for the task at hand.

MARCEL - A very thin wafer of metal between the two friction surfaces. It's got a slight wave to it and acts as a "spring" to make your clutch engagement smoother. It fully flattens under the clamping load of the pressure plate. Interestingly, all of the torque of your engine is transmitted through this thin plate of steel. Pretty amazing if you look at it.

BODY - For lack of a better term, I'm calling the outer part of the clutch that you can readily see the "body". The Marcel is attached to this, and the two sides are riveted together with spacers between them.

HUB - This is the center piece that fits over the splined transmission input shaft. The hub is attached to a plate that is sandwiched between the two sides of the clutch body. On the Mirage, this attachment appears to be insulated with a thin nylon bushing, probably to soak up some vibration.

SPRINGS - The springs go between the hub and the body of the clutch. That springy forward motion you feel as you release the clutch is these springs in action! They absorb some of the torque and then release it. Sort of gives it a little "bouncy" feeling. Without these springs, clutch engagement becomes VERY harsh. That makes starting the car off from a stop very difficult. It also puts a lot of torque INSTANTLY into the driveline, which can be bad for transmission and differential parts. On the plus side, clutch engagement happens VERY fast, and shifts can be a lot quicker. Potentially good for track racing (where the launch isn't important) if you don't mind putting extra wear on your driveline. Pretty horrible for a daily driver. If you ever have the option to choose unspring, don't!

How it all works...

When you release the clutch pedal, the pressure plate clamps the friction surfaces against the flywheel. As the marcel compresses, the clutch engages smoothly, and the body of the clutch disk turns. The body compresses the springs against the hub plate, which further cushions the engagement, and the car starts to move. As the clutch is fully release, the springs normalize and release some of their energy. They continue to absorb some of the variations in torque from the engine, thus not introducing ALL of the engine vibrations to the driveline.

What's going wrong...

I'm going to guess that the springs are either too few, or too soft. If you look at a lot of clutches, you'll see that many of them have 6 springs. Ours has 4. This is likely where they tried to save some weight, not a bad idea. And the springs are probably soft because they're trying to really smooth out the driving experience. It's an economy car, not a race car!

Why do I say this? Let's look...

Here's one of the springs in my clutch.



Notice the grooves in the end of the spring? Bonus points: notice that the grooves are really bad on one end and not so much on the other? One end is the end that takes the load on acceleration, the other under deceleration. Hard launches are killing these springs. Hard downshifts, not so much. (though there is wear on both ends) All four springs show the same wear patterns.

What's causing the grooves?

The grooves are where the spring contacts the hub plate in between the two body sections. Hard to see. You can't even see the wear on the springs until you rotate them to where the worn spots are visible.

It's that silver-colored plate in the middle that's creating the grooves.



Here's a close-up of the most worn spring. It shows something that I didn't even notice before I took the photo. There's a crack in the "body" plate at the corner of the spring window, AND there's a chunk broken out of that same corner! Sure, hard launches and all that... but, come on... the car only has 74 ft/lbs of torque! While I feel the rest of the clutch system (it's torque HOLDING ability) is over-built, I think this whole spring setup is inadequate (for my purposes). The springs are too soft, and/or too few. And the body plates that hold them in are either too thin, or poorly designed. I'm not faulting anybody here. The thing is designed for easy and cheap production and assembly, and designed to be daily driven in a normal fashion... and expected to last 100k miles. It's fine for what it was designed for.



In a nutshell...

My clutch springs and the body material that supports them are wearing. A lot. They already can move as much as 2mm fore/aft, and rattle quite easily. The grooves in them left unchecked, and continuing my current driving habits would eventually cause chunks of the end of the springs to break off. I'm guessing it would happen no later than 80k miles. (but, if we assume that most of this wear happened during MY 10k miles with the car, then it could happen a lot sooner) At that point, the spring action is going to greatly degrade. Things aren't going to feel right, and all sorts of noises could ensue (depending on where those broken pieces end up!). Another thing that could happen is that the body plate openings could wear to the point that a spring actually falls out.

That's about it.

Is this something that the average driver needs to worry about? Likely not. As long as you don't make a habit of making full-throttle quick launches and shifts, you should be okay. Practice a little bit of "mechanical empathy". You can still accelerate at full-throttle, just don't HAMMER the throttle. Roll into it smoothly. Or just enjoy your car in the way that you choose, and expect to replace the clutch at some point.

[Loren]

Alright. Bad clutch disk. Replacing it. I could have taken the clutch disk to a local clutch shop that does custom racing clutches and had them rebuild/improve it. But, instead, I opted to just buy a clutch kit that was readily available on Amazon for $105. It's US made by a company called Southeastern Clutch.

The pressure plate looks exactly the same, but I didn't measure anything. Didn't even weigh it. It's possible that it could be a little heavier or have slightly more clamping force. Or the opposite. I'm trusting that it's at least somewhere close to stock equivalent. Why wouldn't it be?

FWIW, the kit comes with a new throwout bearing, and a clutch alignment tool. No pilot bearing because the car doesn't use a pilot bearing.

Being particularly interested in the clutch disk, I went ahead and put them side-by-side and did some comparative measurements and observation.

DESIGN
Almost exactly the same. No markings on it at all (so it's not a rebranded Exedy or something, they likely make this in-house), but precious few visible differences from the stock clutch.

SPRINGS
Both disks use springs made from 3.6mm thick wire. (there are also secondary smaller springs inside those springs... I didn't measure those) The springs are approximately 190mm diameter (hard to measure that without disassembling the disk), and 37.5mm long.

Knowing the wire thickness and the diameter of the spring, all we need to know next is the number of active coils in the spring to determine the spring rate. I don't care what the actual rate is, so I didn't calculate it. But... The stock springs have 6 active coils and the new springs have 5 active coils. All else being equal (and it is), this means that the new clutch has springs that are a little stiffer than stock!

Score one for Southeastern Clutch! Maybe they used that part because they just happened to already stock it, and it was close enough. Maybe they did some math and decided that they didn't want people complaining that their clutch springs bounce around and break too easily? Can't say why. But, the fact that they are stiffer springs... based on my observations, I think that's a good thing.

The springs on the new disk do rotate fairly easily, but nowhere near as easily as the worn stock clutch. The fore-aft movement of the springs in the worn stock clutch is up to 2mm. The new clutch is barely 1mm, but there is some movement. It's a design thing.

MARCEL
Mostly looks the same. But, if you look at the photo of the stock disk, you'll notice 8 gaps of about 1" length right at the inside edge of the friction surface. That's not present on the new clutch. Does it make it stronger? Stiffer? Heavier? Or is it just easier to make that way? I'm going with probably a little easier to make, and also stronger. Southeastern Clutch does appear to be a "performance clutch" manufacturer, so they've probably seen their share of failed clutches, and have learned how to make them last. (Exedy has even more experience... but, they're also building to Mitsu specs, which may compromise the product in the name of efficiency/lightness or cost or NVH)

FRICTION SURFACE
Nearly the same. Same thickness. I measured it, but forgot to write it down. The OD is 190mm, ID is 133. That's 28.5mm of facing width. The edges of the new disk seem more "square" meaning slightly more surface area. The stock disk has a slight rounded edge on it. (could be another wear-compensation feature giving it more surface area as it wears)

The friction surface has 24 segments molded/milled into it. I think the purpose of those segments is to compensate for wear. As the disk gets thinner, the clamping force exerted by the pressure plate will be less. Increasing the surface area that little bit would help bring the load capacity back to where it should be in spite of less clamping force. That's just a guess.

Interesting difference: The stock disk has a hole drilled in ALL of the 24 segments. But, there are rivets in only 16 of the holes. 8 of the holes are unused. They are there purely for easy of assembly, almost surely automated assembly. No worry about orientation of the friction disk, one less production step. The new disk only has 16 holes. One for every rivet. This probably means that it's hand-assembled. Having 8 less holes in the friction surface, the new disk has more surface area. Enough to matter? Probably not. The clutch already seems to be over-built for the application, so it really doesn't matter.

BODY
One peculiar difference between the two is that the new disk is 1.5mm thicker. The design looks exactly the same, they just used slightly taller spaces between the two body halves. No idea why. I didn't measure the plate thickness... if they're using thicker plates, well, that would be awesome.

HUB
Same. Looks like exactly the same casting. Same thickness, same splines, same diameter.

WEIGHT:
Stock = 1.94 pounds
New = 2.03 pounds

The difference could be the longer spacers used on the new clutch, or simply that the stock clutch is a little worn. It's only a 1.4 oz difference. That ain't much.

[Cobrajet]

I really appreciate the clutch 'autopsy', as I think this may be one of the first times I have seen a Mirage clutch out of the car.

My stock clutch has 106,000 miles on it, and my car is used 100% in-town for delivery. That pedal gets pressed A LOT. It has held up well so far, and I think your assessment of it being adequate for what it is designed for is true.

I am hoping to get to around 120,000 before I have to replace the clutch set.

[stevedmc]

Thank you for sharing this information. I purchased what I believe to be the same kit from eBay back in June.

https://www.ebay.com/itm/CLUTCH-KIT-...torefresh=true

I don't need it right now, so it just sits on the shelf as a spare part until the day my clutch disk finally burns up. At over 220k miles I expect to need a clutch disk any day now, but honestly this cheap car impresses me more and more as I drive it with very little maintenance.

From the looks of it, we have the same kit. Its good to know it is a suitable aftermarket part. I always wondered what those clutch springs were for. Good information. Thanks again.

[Daox]

Great analysis. Love the attention to detail. I agree the clutch is noisey compared to other vehicles I've driven. Now we know why.

[Qrush]

Thanks for the detailed analysis, Loren. I am actually in need of a clutch myself. Unfortunately, I haven't been able to source an aftermarket clutch for a 2014 Mirage yet. I contacted Exedy and AMS Automotive, no response. Both the clutch you showed and the one Steve showed, are for 2015+. Have you a notion why the 2014 is different than the 2015+ clutch?

[Loren]

If there are any clutch differences from 2012-14, I can't imagine why the 2015+ clutch wouldn't at least be compatible. They surely didn't change the spline pattern on the transmission shaft. And they didn't change the bolt pattern on the back of the crankshaft.

Worst case, if you got a 2015 clutch... it might not fit your 2014 flywheel. You just get a new flywheel and you're good.

But, if Mitsu didn't document the fact that they made a change to the clutch in a particular year... they seem pretty good about documenting their improvements. It's probably exactly the same.

[Daox]

The 2014 and 2017 Mirage have the same Mitsubishi flywheel part number (1120A266). The clutch friction disc is also the same part number between the years. So, the clutch will fit. Its just a mistake on the clutch manufacturer's listing.

[2of9]

Awesome in-depth review Loren. Much appreciated. It doesn't look too bad but I guess with that many autocross events, it could be a given with that sort of wear. I definitely agree that if someone is just driving their Mirage around (point a to b), there wouldn't be that much wear on it as well. Really nice to see how well the clutch held up. Good to see that the clutch still holds up. I've driven a car with a 6 puck clutch disk (no springs) and it literally felt like the clutch was on/off, waay difficult for a daily driver but excellent for track use I'd say.