Jump to content

Recommended Posts


The Cosmo is in pretty good condition for a 26 year old car, the only real rust I’ve found so far was in the exhaust back boxes (mufflers if you like).  I noticed some blowing a few years ago and got some patches welded on the boxes as an interim solution to get it through testing, I’ve been planning to replace them since and they just got worse over time.  They originally had an extra thin layer of steel covering the top with a sheet of asbestos-like padding sandwiched between.  I guess this was for sound and heat proofing but any moisture that got in there was trapped and rusted the boxes.

Original lhs back box with patches of Alfa Romeo added


OEM replacements don’t seem to exist anymore so another solution was needed.

As much as I’ve drooled over the Kakimoto system that’s available for it in Japan, it would take from the originality of the car, drain my budget (about €1600 before shipping), and leave me needing supporting mods to allow for the extra flow and reduced backpressure.

A few 2nd hand sets have come up, not cheap, without knowing what condition they were like on the inside, I wasn’t willing to risk a lot of money on a set and any that were cheap enough to be worth a gamble, the sellers fell off the face of the planet after I enquired.

Another option was to get somebody local to fit magnaflow boxes or similar but again, no way of knowing how flow would be affected.

So a while back I decided to cut open one of mine so I could figure out how they work, it was beyond repair anyway as some of the insides had come loose and were rattling around in there.

Guts removed (rhs box)



An idea of the pipe routing



Edited by gl
Link to post
Share on other sites

The 20b has 2 back boxes with twin outlet pipes on each.  The outer pipe on each side has a butterfly valve which is opened by vacuum at high load.  I wanted to retain this feature if I could.

Upon removing a section from the shell of the box, the box’s complicated guts were revealed.  It took a lot of head scratching to figure out the direction of flow and what went where. 

Each box has 2 internal baffle plates which hold the pipework in place and also divide the box into chambers.  The inflow pipe carries gasses through chamber 1 and releases  them into chambers 2 and 3.  A hole in the first baffle allows the gasses back into chamber 1 where they can only go into the smaller pipe.  The smaller pipe does 2 x 180degree turns before flowing out the exit.  The bigger pipe has the valve which opens by vacuum and allows gasses from chamber 2 to exit more directly.

 I made a few sketches and drew it up in cad. 


View of baffle plates with holes needed



4 for the left, 4 for the right



Side view showing gas flow



Plan view showing gas flow (rhs box)


Confusing isn't it?


Edited by gl
Link to post
Share on other sites

At this stage I started to see it as a collection of pieces that needed to be made and joined together so I priced up the materials to make them with t304 stainless and it seemed achievable.

There were a few issues with making exact copies in t304.  I couldn’t get mandrel bends with a small enough radius so had to adjust the design to suit the minimum radius available.  Calculations showed that the extra radius would have minimal effect on back pressure (<2mBar difference).  I adjusted the baffle plates to suit and sent off my dwg to a company who cut them out for me. 


 I also ordered all pipes and flanges needed and enough sheet steel to wrap the 2 boxes, all plates, sheet and pipes were 2mm t304 stainless.

I made a wooden jig to aid with assembly and used it initially to line up the baffle plates, then to test fit the pipes.


Mazda had done something in the original exhaust which helped me a lot in this.  Although the boxes were made from mild steel, the outer pipe with the butterfly valve was made from stainless.  This meant I could reuse these pipes on the new system and not have to transfer the butterfly valve into a new pipe.



Edited by gl
Link to post
Share on other sites

In the meantime I had been on the hunt for somebody who could weld the boxes up for me.  A name came highly recommended and I brought all my bits to him.  He made his own jig and welded all the bits together for me.  Fast forward a considerable amount of time and this was done, mockup no.1 ready for inspection.  Some extra steel was removed from baffle no.3 to aid flow between sections 2 and 3.  The sheet I had supplied to wrap the boxes was rolled into shape and tacked.

The rhs box and it's skin



Soon after that, this was ready, the right hand box ready for a test fit



 I decided to re-use the hanger brackets from the original exhaust, I’d like to have replaced them with stainless but It involved a lot of extra work which can be put off until a later stage.  These were just tacked on for test fitting.  A few small adjustments needed to be made after the first test fit just to get everything lining up right.


Soon there were 2



Edited by gl
Link to post
Share on other sites

Nice 1 Nico

Just to add I've attached the cad file here which has everything drawn to scale, this is what I sent the company to cut out the baffles and can be used to measure pipe lengths etc.

One of the baffles in each box (rb3 and lb3) needed an extra hole cut to allow airflow between chambers 2 and 3, this could have been done at the cutting stage.

Pipes used were 45mm and 57mm.



Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Create New...