Mazda MX-6 Forum banner

1 - 19 of 19 Posts

·
Premium Member
Joined
·
2,003 Posts
Discussion Starter #1
I finally got around designing the exhaust manifold for the engine. With the new job I got to use some new software which turned out to be quite decent for manifold design.

The requirements for the exhaust are quite quite demanding. The exhaust had to be compact enought to allow both turbo and supercharger installation, it had to run the exhaust primaries for proper twin-scroll pattern, and I could only use fixed radius bends (40mm OD 60mm radius tube bends).

Turns out the first arrangement I thought was entirely unfeasible, and had I tried that with actual pieces I would've only gotten an expensive lesson.... After tweaking the design a bit, I ended up with this:



Now, at first it seemed like an okay deal, being symmetric it would offer equal length runners for each primary pairs. However, the collector would've needed a corkscrew inner walls which I was not happy about, and the design was too wide, it brought exhaust manifold right next to the cam belt, not a happy situation. It also constricted access to space below exhaust headers. Also the port spacing was wrong but that was some of the least difficult parts to fix. I started with the paths for runners 1 and 3 and redid much of the routing and ended up with this:



Now, it's taller on the transmission side, but puts very little heat load on the compressor side (the turbo will be reversed to normal 626/MX6 arrangement; compressor side on the right side, turbine on the left), also completely avoids bringing the runners to the cam belt vicinity and arranges the runners in proper order without ugly trick collectors. It also lines up the compressor pretty neatly for an A2W intercooler on the way to the supercharger. There should be just enough room for all this without having to go for frame-forward radiator. It's not an equal length exhaust, but it's not as critical as with NA, fortunately.

Also, in contrast to the first design, the runner inlets are this time flared to 50mm and not squashed to match the ports in the head. This is to spread the loads on a wider area, but also has some interesting implications. according to literature anti-reversion devices are recommended when dealing with non-equal length runners, and as close to the head ports as possible; thus, here's a detail in the design:



The pockets are to increase anti-reversion effect from the inherent stepped port setup, ideally the intention is the pressure pulse is forced to return to exhaust runner instead of allowing radial flow. Don't know if it makes any difference, but it looks cool... The design of the flange is an attempt to combine the benefits of thin flange (quick heat dissipation, lightness, less abrupt wall thickness change) with those of a thick flange (stiffness, strength). The round bosses in the design are intended for turbo support brackets, possibly threaded holes for rod ends.
 

·
Premium Member
Joined
·
3,355 Posts
Yes - Brilliant!

The only thing I would say there is with no compromises like packaging (space) You can knock out things like that and for out and out maximum HP there is nothing better.

Most of us have to be more pragmatic. Log manifolds are horrible from a flow perspective, but brilliant for packaging.

I've often wondered why manufacturers close the exhaust valves early on turbo motors - maybe to deal with short and non ideal runner length.

The hardest part is finding the optimal design between the ideal and the practical reality.

I've thought about going to twin (small) turbos and completely splitting the runners to feed them 2/3 and 1/4. The good thing about that is that there is no exhaust overlap then!
For packaging I would make one a high mount and the other a low mount.

In reality I am still at the point where I'd be slicing off an EVO4-8 flange and welding on an FE DOHC one.

The design tool is pretty cool though, I wish I had access to it.
 

·
Premium Member
Joined
·
2,003 Posts
Discussion Starter #5
Thanks, all! :) The software is Moment of Inspiration 2.0, ironically one software I could've easily bought for myself had I known about it before. It's only $300. Flanges are carryover from earlier times, done in CATIA, SolidWorks and recently in Geomagic Design Elements (former Alibre). MoI doesn't have any specific pipeline design capabilities, but it is very flexible and quick to use, and allows quick 3D sketching.

What's really interesting is how to go on about building the thing. There is a shop offering precision bent 309S heat resistant stainless steel tube bends, but the total cost will be about 750 USD...
 

·
Premium Member
Joined
·
2,003 Posts
Discussion Starter #6
Uh-oh, another missed estimation, digital mock-up doesn't look good:



...talk about a hood bulge. "Is that a truck turbo or are you just really happy to see me". Okay, talked about. I think I need to try and depress the collector by at least 50mm to make it more acceptable.
 

·
Premium Member
Joined
·
3,355 Posts
More, so as to allow for enough heat shielding so as not to burn paint off the hood (one reason I am not a fan of Hi-mounts)
 

·
Premium Member
Joined
·
2,003 Posts
Discussion Starter #8
I'm not above making a hood bulge if need be, but I think I'll be doing an Evo-style vent on the hood instead, it'll improve engine bay cooling significantly (unlike windshield base gap which phucks it up totally) and would be right over the turbo. Especially if underbody shield/fairing is to be added later, the stock cooling arrangement would become very inefficient.

The engine placement and the car silhouette aren't exact, so it's all very rudimentary at the moment, but it'll tell we are going to run into clearance issues... However, it turns out I had inadvertly moved the exhaust/turbo combination up by 20mm, so that'll help right off the bat; shortening the runners by 35mm only requires redoing one tube, and the collector could allow further 20mm reduction, so we're already 75mm lower, which is pretty much all that's needed.


(invisible turbo flange, the hottest new thing)


CAD accuracy is not required nor attainable in real world, so allowing a bit of mismatch that's hammered into shape will allow a bit more adjustment. It would just be nice to have the design fully viable without having to manually adjust it.

Oh, and it seems the exhaust might even allow A/C compressor installation, which I would consider a miracle of biblical proportions.
 

·
Premium Member
Joined
·
2,003 Posts
Discussion Starter #9
Revision #3. Since keeping the turbo centered between the exhaust port center gave no benefits, I moved the turbo between cylinders 3 and 4 for more hood clearance (on more likely, less hood protruberance). At the same time the the difference in length between runners was cut in half, more or less! The manifold is taller than ever before, but it sort of follows the shape of original exhaust manifold, so it could actually be a good thing for heat management. I got to measure an untouched engine bay and the space is *rrreally* tight.

 

·
Premium Member
Joined
·
1,257 Posts
This is awesome i wish i knew how to use these programs better. Tried solidworks but didnt make it very far. What if you lower the turbo flange below the head flange in ur designs...that would help with hood clearance but im unsure if it would cause any other negative benefits
 

·
Premium Member
Joined
·
2,003 Posts
Discussion Starter #11
Hello Gabe! unfortunately lowering the turbo flange is not an option, I spent a good amount of time trying to do that until I realized the compressor shroud wouldn't fit:



The compressor shroud is tangential (or close to) with the exhaust intake flange face. Thus the turbo flange must be an outer extremity, or the compressor will come too close to the exhaust tubes. Angling the collector towards the engine might give a slight drop though and certainly improves radiator clearance which also is a major issue.

BTW I didn't get very far with Solidworks either, the basic 3D sketching just didn't work out for me in this kind of design work. There is a pipeline design tool too, perhaps if I had that I could make five of these designs a day. I don't so I can't, unfortunately. I'd love a software where you could grab and manipulate tubing elements directly and everything would simply adapt.

Unless I can figure out still a better shape, I'll start doing drawings for this one as soon as I've figured out the collector and wastegate ports.
 

·
Registered
Joined
·
46 Posts
Wow. It's nice to see the proper through process going in to something. :tup:

Is there a reason you are set on top mount? Is that specifically for supercharger clearance/better packaging, as you said in reference to your intercooler?

I'm going to guess that's a divided T4 flange? What turbo are you using if you don't mind me asking?

Oh yeah, and where do I sign up for one of these when you have them made?;)
 

·
Premium Member
Joined
·
2,003 Posts
Discussion Starter #13
I don't mind questions at all :)

The top mount is indeed for suprcharger clearance, I'll have turbo feeding the supercharger through a water intercooler.

At first I planned on reversed setup, supercharger on top and turbo at the bottom, but the heat dissipation was a problem, the belt arrangement became messy and the supercharger bracket would've been rather unwieldy as well. Technically both did fit at low mount position, but ducting didn't.

Had I went for a RWD like I thought at one point, I would've put the supercharger on the alternator side, but in GD that wouldn't work.

It's a split T4 flange, that's correct. The turbo is Holset HX50 MFS, 19cm turbine shroud, and the supercharger is Eaton "M65" (M45 in a fancy dress). Running the M45 at 1:2 overdrive ratio should give a decent pressure ratio of 1.5, and the maximum revs are nicely in line with the 94mm stroke.
 

·
Premium Member
Joined
·
2,003 Posts
Discussion Starter #15
I found my old DOHC HEAD cad project! Envelopes are the way to go while modeling, but nothing really beats seeing the actual parts coming together, so here goes....



It's a *tight* fit! and a lot of work. Lots and lots of work. The supercharger is missing the intake part and upper half, and there's the secong charge cooler to install (in this build, inter- and aftercooler hold more than symbolic difference!)
 

·
Premium Member
Joined
·
2,003 Posts
Discussion Starter #16
Okay, I got the supercharger assembled... and it's not looking good fitment-wise. The photos hardly tell the whole story, but here's a few shots.




The bypass duct reaches to spot not farther than 100mm forward and 120mm below the exhaust port centerline on the head. It's possibly to create the clearance needed by bunching the most of the exhaust piping to the left side of the engine. It's going to be an incredibly tight fit. Since the exhaust will be about $800 per for just the materials, it's pretty important to get it right the first time.

However, there was a fortunate accident too. I bought a chinese 65mm throttle body for supercharger bypass valve. As it happens, the upper half fits very snugly inside the throttle plate, and the vertical clearance is a bit under 1mm. Just a bit of trimming the base and it'll look as if made to fit. Here's the photo:

 

·
Premium Member
Joined
·
3,355 Posts
That thing is a monster.
What is that square "box" that sits past the front of the block ?

If it were me i'd ditch the Eaton and go with an SC14 or something like that. I had one several years ago to see if I could work out somewhere it could fit and decided it was all too hard and sold it to one of my work mates to put on his Fiat 124s.

Alternately maybe its time for some diagonal thinking, what about an extension shaft for the drive and placing the SC over the top of the gearbox. I've seen that trick use on a KL powered GE.
 

·
Premium Member
Joined
·
2,003 Posts
Discussion Starter #18
The aquare box thingy is an induction sound silencer, to mute the supercharger howl. Benz item and all...

SC14 (and TX15) are problematic with the both ports being perpendicular to the supercharger body. The bottom port would face the subframe rail and require a much higher mounting location. An actual M45 would be ideal (shorter, less tall), but having spent time, money and effort on the M65 already I'm going to see that process through (and the M45 seem to be a bit pricies than M65s too)

Moving the supercharger over to the transmission side could be an answer, but then again, the new gearbox has quite a lot of stuff going on in the same area. Also, I'm a bit uncomfortable with the idea of an exposed shaft spinning 14 000 rpm in the engine bay.

Compromises and concessions aplenty. Engineering in it's true form.
 

·
Premium Member
Joined
·
2,003 Posts
Discussion Starter #19
Unemployed again, since August 2014... it's not a good moment to keep pouring money on a car build (and I have two of them).

Anyway, since I have no problem with free time, I've finished (hopefully) the design on the exhaust. Here's the beast:



Wastegate runners required some more modifications to the design, and while the bounding box hasn't changed much, there are important changes where the exhaust runners are placed in regards of the supercharger installation The WG runners are to be made of 304 stainless steel, a bit lower grade than the actual exhaust runners (which are 309), and they're of smaller diameter and tighter bends too. I'm happy the way they turned out, they leave ample room for the downpipe and do not block exhaust studs or turbo support bracket. The problem is when you have only one bend radius (and relatively large at that) to work with, a seemingly innocuous change is likely to alter the routing so much it'll interfere with another runner. The change will propagate through the entire design, often ending up in an unworkable or less dense routing.

Now, I'm really cash-strapped at the moment, but hopefully I can scrounge something up to get the exhaust made sooner than later. it's the last major part in the engine build.
 
1 - 19 of 19 Posts
Top