Cast iron 4:2 exhaust manifolds

Thought about making a new thread to show the work on my fabricated downpipe but a quick read through this one focuses on the downpipe for much of it as well. Here is a quick update on my progress so far.

The 4-1 manifold I have is one that is angled slightly to the left (American driver's side when installed). The flange was machined my a machinist after hours for $60 dollars total using 1/2" 304SS. While its a bit of overkill on thickness the stock was free. Purchasing 3/8" x 4" x 6" which is what would be needed to match the manifold flange was another $40 after shipping :eek:. I think I'll take the few extra grams to the car.

I wanted to add some flex to the system as I've seen many people complain about weld zone cracking due to fatigue in their performance manifolds. At the top are two bellows joints and the bottom a braided 6" flex joint. I wanted the bellows just below the flange but could not as they are too large in diameter and would have to be offset. With that, I incorporated the 90 degree rotation first to better align the duel pipes out the bottom. Slight adjustments needed to better align the two but
IMG-4390.JPG

The bellows are directly below that. From these photos slight tweaks will be made before tacking this week to better align them parallel with the rear firewall and frame. I've also incorporated the 2-1 merger with the 90 deg elbows to hopefully allow only 1 pressure wave reflection back to the next valve. Both, mergers and hard bends send pressure waves back towards the engine. According to my local expert on exhaust tuning at a targeted 5500 RPM performance benefit cylinders 180 degrees apart should have a separation of 24-30 inches. Quick measurements had the end of the manifold to be 14 inches and from there to below the frame rail to be another 12 inches putting me at 26" give or take. This pic shows a little long as I didn't want cut the elbow merger until needed.
IMG-4394.JPG

Once below the frame its straight rearward a few inches then 90 degrees towards the threaded flex joint and new stainless V-band clamp on the otherwise stock muffler. More to come as I try to tack some pieces together tonight to confirm positioning before fully welding.
 
Nice work AK. I like the orientation of the bellows/bends, as you say they are too fat to fit side by side otherwise.


AKimball92 said:
According to my local expert on exhaust tuning at a targeted 5500 RPM performance benefit cylinders 180 degrees apart should have a separation of 24-30 inches.
Click to expand...
I've been struggling with the dimensions a little. According to the formulas, the length of the primary tubes (the tubes from the valves/head ports to the first merge {four into two}, essentially the exhaust manifold) are too short with these manifolds. So that throws off all the other lengths (secondary and tertiary). I'm actually ending up struggling to make everything short enough, as odd as that sounds. And I'm aiming for more torque gain than top end. I've double checked the math but the real issue is the design of the cast manifold is off for an ideal set up. I guess this is where a true tubular header might benefit. However I'm not that concerned about it; if I don't get that extra 1 or 2% gain, the loss of 1 horse power isn't going to kill things.

I don't recall if you already stated this, but based on your description above, it sounds like you will be going under the rear-suspension cross-member rather than through it?
 
Dr.Jeff said:
Nice work AK. I like the orientation of the bellows/bends, as you say they are too fat to fit side by side otherwise.
Click to expand...

I had to rotate it to get the 2-1 merge and elbow how I originally planned. I wanted the bellows as close to the engine as possible. I cannot technically state why, I just through having the joint closer would allow more movement. Also, having them staggered would cause a weird pivot plane although that too probably wouldnt matter.


Dr.Jeff said:
I've been struggling with the dimensions a little. According to the formulas, the length of the primary tubes (the tubes from the valves/head ports to the first merge {four into two}, essentially the exhaust manifold) are too short with these manifolds. So that throws off all the other lengths (secondary and tertiary). I'm actually ending up struggling to make everything short enough, as odd as that sounds. And I'm aiming for more torque gain than top end. I've double checked the math but the real issue is the design of the cast manifold is off for an ideal set up. I guess this is where a true tubular header might benefit. However I'm not that concerned about it; if I don't get that extra 1 or 2% gain, the loss of 1 horse power isn't going to kill things.
Click to expand...

I will admit, I haven't done the calculations on my own either. I have a Matlab code started but there are some constants I cannot get good data for. I think these would cause enough variation to make any reasonable lengths correct or false. I decided to trust the technical specialist at my employer (many might know where that is) and a friend of my Internal Combustion Engines professor at college. He also taught the lecture dealing with the exact coding I am trying to work on. Once I got his answer, it was less of a priority that I complete the code.

Dr.Jeff said:
I don't recall if you already stated this, but based on your description above, it sounds like you will be going under the rear-suspension cross-member rather than through it?
Click to expand...

Yes, I wanted to keep the cat. but to make the secondaries long enough and to follow the original routing through the firewall would have doubled the amount of elbows I needed to by, not counting any mistakes, and the amount of welding needed.
 
For the tube lengths' calculations I went off two sources:

One was the work of a fellow member here a few years back. You might recall the thread where we were developing the "ideal header" (all tubular exhaust header) to make from scratch. He did a ton of research and found all the reference he could on how to calculate everything for the specific specs of the engine, desired power band, etc. Those dimensions were calculated to build the header and corresponding exhaust system.

The other source I used was a very old book by Colin Campbell, "The Sports Car Engine, Its Tuning and Modification". It's considered to be one of the original 'go-to' references for performance modifications on small engines like the X's. Funny side note, it was the first hardback book I ever bought new from a bookstore...I think I was about 7 or 8 years old (what does that say). Anyway this was written looong before the computer was even created, so it gives an entire chapter of theory and mathematical calculations on how to design an exhaust system.

Both sources focused on 4:2:1 designs, and both had very similar results (I'd have to look up the details again if you are interested). That's what I was going by when I said the stock manifold has runners that are too short for an ideal design (based on my parameters). So I'd have to go back and recalculate everything to accommodate the shorted "primary" lengths on these manifolds (haven't done that). To be honest I don't think it will make that much difference; there are other practical considerations to be taken into account, plus any given design is made to optimize performance at a specific RPM and we don't drive normal cars at one RPM constantly.

Interestingly I found that getting the desired lengths for the downpipe and exhaust (the secondary and tertiary pipes) was a lot easier than I thought it would be for a mid-engine car with little space behind it. In fact the calculated lengths are almost too short for some layout options. Therefore I choose to go above the rear-suspension cross-member. This layout was actually much easier for my design; but I'm not running a cat, I am not trying to keep anything stockish, and my car will be VERY low (very little ground clearance) and it keeps the pipes up away from the ground.
Obviously there is more room from side to side than from front to back, so the muffler needs to be mounted transverse. That means a couple 180 degree bends of the exhaust to accommodate it, and that puts a lot of length into that tertiary section.

To minimize the number of bends I choose to keep the secondary pipes stacked front-to-back (rather than twisting to side-by-side) as they travel down and through the opening above that cross-member. That allows them to bend to one side just before the final merge (2:1). However I am not adding the flex joints at the flange like you have, so that made it much easier. It is definitely preferred to have those flex joints, but this car will be more of a "track day" build. So minimal exhaust (little weight) and minimal use (little wear/tear like a road car has), plus vibration/comfort isn't a real concern.
 
Dr.Jeff said:
For the tube lengths' calculations I went off two sources:

One was the work of a fellow member here a few years back. You might recall the thread where we were developing the "ideal header" (all tubular exhaust header) to make from scratch. He did a ton of research and found all the reference he could on how to calculate everything for the specific specs of the engine, desired power band, etc. Those dimensions were calculated to build the header and corresponding exhaust system.

The other source I used was a very old book by Colin Campbell, "The Sports Car Engine, Its Tuning and Modification". It's considered to be one of the original 'go-to' references for performance modifications on small engines like the X's. Funny side note, it was the first hardback book I ever bought new from a bookstore...I think I was about 7 or 8 years old (what does that say). Anyway this was written looong before the computer was even created, so it gives an entire chapter of theory and mathematical calculations on how to design an exhaust system.

Both sources focused on 4:2:1 designs, and both had very similar results (I'd have to look up the details again if you are interested). That's what I was going by when I said the stock manifold has runners that are too short for an ideal design (based on my parameters). So I'd have to go back and recalculate everything to accommodate the shorted "primary" lengths on these manifolds (haven't done that). To be honest I don't think it will make that much difference; there are other practical considerations to be taken into account, plus any given design is made to optimize performance at a specific RPM and we don't drive normal cars at one RPM constantly.

Interestingly I found that getting the desired lengths for the downpipe and exhaust (the secondary and tertiary pipes) was a lot easier than I thought it would be for a mid-engine car with little space behind it. In fact the calculated lengths are almost too short for some layout options. Therefore I choose to go above the rear-suspension cross-member. This layout was actually much easier for my design; but I'm not running a cat, I am not trying to keep anything stockish, and my car will be VERY low (very little ground clearance) and it keeps the pipes up away from the ground.
Obviously there is more room from side to side than from front to back, so the muffler needs to be mounted transverse. That means a couple 180 degree bends of the exhaust to accommodate it, and that puts a lot of length into that tertiary section.

To minimize the number of bends I choose to keep the secondary pipes stacked front-to-back (rather than twisting to side-by-side) as they travel down and through the opening above that cross-member. That allows them to bend to one side just before the final merge (2:1). However I am not adding the flex joints at the flange like you have, so that made it much easier. It is definitely preferred to have those flex joints, but this car will be more of a "track day" build. So minimal exhaust (little weight) and minimal use (little wear/tear like a road car has), plus vibration/comfort isn't a real concern.
Click to expand...

Do you have any pictures of drawings of your set up? Could you share some of your calculations somehow as well? Maybe we could set up a plug and play for people to tune their own exhausts. No GUIs or anything, just let them find their own calculators. If I wanted to suffer through the extra few bends and cuts I would like to continue with the Cat. If anything I could always straight pipe it if I wanted to be a rebel, cheap, or feel more power.
 
For starters here is the old thread (or one of them) with the background on calculating the dimensions:
https://xwebforums.com/forum/index.php?threads/my-ideal-header.16768/

And here are a couple other decent exhaust discussions that might help:
https://xwebforums.com/forum/index.php?threads/exhaust.31843/#navigation

https://xwebforums.com/forum/index.php?threads/mocking-up-an-exhaust-system-done.23063/

I will have to look for the dimensions I came up with (although I might have posted them previously?). In the end I am using actual tubes that don't strictly fit the calculations. Mostly because I already have a bunch of similar materials and this is a very low budget build. Likewise, as mentioned the lengths will differ some. For my goals the calculated dimensions are more of a general guide than actual build plan. I hope as an engineer you are more disciplined than I am. :)

As for my specific design, so far I've done the basic mock-up on the car with various mandrel bends in my hands to figure out how I want it. So no actual pieces to show yet, sorry. I'm old school and work out of my head and simple paper sketches, so again nothing I can post. But I will when things are actually getting built (although my turbo project has put this one on hold for a while). Maybe I can set some of the pre-bent pieces on a table to try and illustrate the concept, I'll see.

Sorry I can't be more help at this time, I know you are at that point in your build and I wish I had more to show. But this is for the second of two X builds I'm doing, plus a show-car VW that has been ongoing for some time, and three custom motorcycle projects. So this one keeps getting pushed back.
 
From that first link, if those are your prefered dimensions, that appears very doable. In fact, I think my calculations and my friends superior judgement greatly match the numbers provided:
"4-2-1 system,
First pipes (from head) 1.25" O.D., and 15" long,
Second pipes, 1.65" O.D., and 18.8" long (this includes length of the first collector),
Third pipe (to entrance of muffler?) 1.8" O.D., 36.8" long.
All collectors would taper at 7-10 degrees.
...
The total length of the first and second pipes and the collector would total 33.8" "
​
while I never calculated the tertiary length as I deemed that not a variable I could or wanted to play with. But 15 inches for primaries? my rough tape measure calculations had the primaries at 13-14. At least I measured the end of the manifold to be 14 inches. You had the primary+secondary length totaling 33.8" with the comment to shorten the primary if more mid-range power was preferred. I was after a 5500 rpm peak improvement from the exhaust with the length suggestion being 24-30". Those values seem to correlate.

Also another matching characteristic between our exhausts would be the diameters. I chose 1.5" 16 ga. and 1.75" 16 ga. both of those are very close to your diameters. I am not sure if those were set as variables to optimize or limited parameters given the cylinder head limitations. That's how I viewed it, plus the plan to use the 4-2 manifold.

As for the pictures, they were just for curiosity's sake. Had it been 3 months ago, I would have wanted them more. No, now I'm too far down this path to want to go through the firewall.
 
My numbers were almost the same as the ones in that older post. Basically it sounds like we all seem to come to the same conclusions. That's good to know.

The primary tube diameters were based on performance parameters, not head port size, because it was intended for a tube type header (i.e. no cast manifold). But I think they are about the same diameter as the cast manifold if I recall correctly. The diameters I will use for the secondary and tertiary tubes are a bit larger than the calculated ones (like 1/8" larger dia) only because I already had them.

The cast manifold runner lengths vary considerably from port to port due to its design. I think they averaged something like 11" (off the top of my head)? Which to me seem considerably shorter than the 15" calculated. I don't recall how this might be corrected for with the lengths of the secondary or tertiary tubes? For example with the primaries a little shorter, can the secondaries be adjusted a little longer or shorter to compensate and keep the overall goals about the same? It's been too long since I studied all this to remember.

The length of the tertiary section basically equals: the calculated length for the overall system (all sections combined) less the lengths of the first two portions. In other words, whatever length is needed to achieve the intended total length, from the valves to the tailpipe tip. So it really isn't a particular dimension on its own, but a means to get a total system length.
 
Not sure if this gives you the intended plan, but here's a couple photos of how the downpipe will be designed.

From the exhaust manifold the first two pipes will remain parallel front to back. In this pic there is only one pipe, so imagine two siamesed, and matching the locations of the manifold's outlets (couldn't keep it all lined up for the pic):

021.JPG


They make a 45 degree bend through the opening above the rear cross-member, still stacked together front/back (like above). Then after crossing that opening, they make a 90 degree bend to the driver's side, at which point they become on top of one another (top to bottom) rather than front to back:

020.JPG

022.JPG


From there I originally planned to have both pipes travel to the left side and make 180 degree bends before merging into 1 pipe. But that proved to be too long of runs. To get the preferred secondary lengths, they will merge before making the 180 degree bend:

024.JPG


The muffler travels transverse to the right side. Then a final 90 degree bend heads straight to the exhaust notch in the rear valance panel (at the stock opening). Hope this makes a little sense. But you can see that by going through that panel above the cross member the tubes are relatively simple without a lot of bends.

And I forgot to include in the pictures the flex joint. As stated earlier there won't be two at the manifold flange unfortunately (not enough room with the pipes laying front to back). But one larger flex will be right at that first merge point (immediately after the pair of 90's, at the red arrow above) where the tertiary begins. At least that allows movement between the downpipes and the muffler section.
 
Forgot to mention. While I had the manifold out for the photos, I did a couple quick measurements.

Port inlets are 28mm (~ 1-1/8"), so smaller than ideal. The twin outlets (other end of the manifold) are 35mm (~ 1-3/8"), so also a little small but close. And the runner lengths are roughly 9" for the middle two and 10" for the outer two, so average is less than 10" long. I think that's 33% shorter than the calculated primary length of 15". So the cast manifolds are not really very good after all, but it's as close as we have without going to a tubular header. These dimensions are from the European version (7691880 I believe) manifold. It has a fairly central outlet position, slightly off to the left (driver's side). How do those dimensions compare to the one you are using?
 
Last edited:
Dr.Jeff said:
Forgot to mention. While I had the manifold out for the photos, I did a couple quick measurements.

Port inlets are 28mm (~ 1-1/8"), so smaller than ideal. The twin outlets (other end of the manifold) are 35mm (~ 1-3/8"), so also a little small but close. And the runner lengths are roughly 9" for the middle two and 10" for the outer two, so average is less than 10" long. I think that's 33% shorter than the calculated primary length of 15". So the cast manifolds are not really very good after all, but it's as close as we have without going to a tubular header. These dimensions are from the European version (7691880 I believe) manifold. It has a fairly central outlet position, slightly off to the left (driver's side). How do those dimensions compare to the one you are using?
Click to expand...

You have the same manifold as I do. I didn't measure my outlets to be 35mm but at 33. I did however port them out a bit to make the holes round and to match the downpipe I.D. Off the top of my head I don't know the new diameter but will check when I get back to work tonight. I will also remeasure the primary lengths and see what i come up with.

To get through the factory hole above the cross member for me required a decent turn to the RHS of the car. My factory hole is about 4 inches to the right of where my downpipe is coming out, between the cross member and control arm mount. Make sure you can get that angle before you go too deep into this.

Another though on this is that each decently sharp bend sends another pressure wave back up the exhaust pipe. I am not sure if you will get the desired scavenging with the 2 sharp bends before your 2-1 collector.
 
The outlets were not round as you say, so the measurement posted was the largest point. I guess the actual diameter is more like 33 or 34mm, hard to say exactly. But like you, I will need to port it regardless, so it really doesn't matter. And all of that is also true for the inlet diameters. Actually they are a distinct oval shape on the head. Again, porting will be needed.

I think I mentioned somewhere that on mine the panel above the rear cross-member has been cut out. So there is one large opening rather than the smaller one that is offset to the right. Therefore I am able to aim the pipe straight rearward with no extra bends.

Agree about every bend adding pressure waves and resistance. But that seems the only way to get any type of muffler longer than about 10 or 12 inches to fit. If the muffler isn't positioned transverse, that's about all you could use. Unless you have a different design idea? I am most certainly open to a better way to do it.

One advantage to routing the pipes over the rear cross-member is there aren't any bends needed to go down, under, then back up around it. The manifold's outlet is above the opening, allowing the pipes to have a 45 degree angle from the manifold straight through the opening. So I think it actually has fewer direction changes this way.

I thought about running the pipe straight from the opening to the rear panel (basically a straight-pipe) and using a "Super Trap" type diffuser as the tip. That way there is only one pressure wave, at the end of the pipe. Also the level of resistance (and sound) can be adjusted with the number of discs added. We ran these on our VW powered dune-buggy's a long time ago. Surprising how well they muffle the exhaust. The drawback to this design is the tertiary length is extremely short, making it difficult to get the total length that's desired. So any advantage might be negated.
 
Andrew, I think we are on the same build plan. Not only with the exhaust but also the intake. I saw your post in another thread; I'm working on a single Weber DCNF set-up. ;)
 
Dr.Jeff said:
I think I mentioned somewhere that on mine the panel above the rear cross-member has been cut out. So there is one large opening rather than the smaller one that is offset to the right. Therefore I am able to aim the pipe straight rearward with no extra bends.

Agree about every bend adding pressure waves and resistance. But that seems the only way to get any type of muffler longer than about 10 or 12 inches to fit. If the muffler isn't positioned transverse, that's about all you could use. Unless you have a different design idea? I am most certainly open to a better way to do it.
Click to expand...

A real easy solution......would be to just use a muffler like this one from a Saab 900. A transverse muffler but with the inlet that is in the side of the muffer. Lots of cars used a muffler like this. Later Miatas, Kia Sportage, Nissan Murano, etc. Would require just a couple of gentle bends to get to it from the exhaust manifold. Hence very little restrictions. Might not end up with the perfect theoretical lengths of pipes, but something perhaps to consider....
saab 900 muffler 001.JPG
 
rx1900 said:
A real easy solution......would be to just use a muffler like this one from a Saab 900. A transverse muffler but with the inlet that is in the side of the muffer. Lots of cars used a muffler like this. Later Miatas, Kia Sportage, Nissan Murano, etc. Would require just a couple of gentle bends to get to it from the exhaust manifold. Hence very little restrictions. Might not end up with the perfect theoretical lengths of pipes, but something perhaps to consider....View attachment 13036
Click to expand...
Can it be used in opposite direction? :D
Then inlet would be in the middle..
 
You must log in or register to reply here.
Back
Top