• By
  • On 23rd March 2015

“What mufflers should you buy?” Pea shooters, megaphones, Dunstall style, British standard, fishtail, cigar type, grenades, shorties, clip-on’s or slip-on’s? As with most things today, the variety of available aftermarket exhaust components is overwhelming, indeed.   The choice will depend primarily on looks and then perhaps fit, but performance and sound are probably just as important (and perhaps even more so).  In this post, I dissect a 1977 HONDA CB550F (super sport) exhaust system in the hope that some of the knowledge gained from looking at what’s inside may be useful in determining what aftermarket mufflers to choose and what may need to be done to modify them for a correct engine match.

Much has been written about exhaust theory in terms of header pipe diameter, pulse flow, gas scavenging, principles and types of sound dampening, and, of course, the all-too-often misinterpreted term “back pressure”. It’s all great stuff, very useful and I highly recommend reading up on it.  But what has not been discussed much and for which there’s very little available information and advice is how to match your new aftermarket exhaust system to your bike’s particular engine.  After all, there are 50 cc bikes and there are 1,500 cc ones.  Some bikes can run at 20,000 RPM, others could barely do 6,000.  And some use carburetors while others are fuel injected and feature engine management that is fully computerized.  Surely, there could be no single one-size-fits-all universal muffler.

The challenge I faced was this:  I was going to switch from a 4 into 1 exhaust system to a 4 into 4.  Why?  See the DESIGN PRINCIPLES post for the answer.  I was going to go from a single original muffler to four aftermarket ones.  Will that have an impact?  What  do I do to ensure whatever I use is good for the engine?  Is there a principle, “rule of thumb”, I should follow?

I decided that one way to get there is to try to keep the “flow dynamics” as close to the original as possible.  And for that, I needed to look inside.

Here is the original 1977 HONDA CB550F exhaust system:


The header pipes are 1.25” (3.2 cm) outside diameter and approx. 1.0” (2.5 cm) inside diameter.  They taper slightly as they terminate into the collector, which has an outlet of approx. 2.0” (5.1 cm) in diameter.

The muffler itself consists of two components: the outside housing and the silencer inside.

005 cropped


The silencer itself is constructed of two pipes that are welded together and are approximately 7” (17.8 cm) long and 1.375” (3.5 cm) in diameter.  The other ends are “flattened” and half circle in shape . The exhaust gasses enter through one of the round ends (inlet), go through the pipe, come out of the half circle end, bounce back to the other half circle end and finally come out through the opposite round end (outlet).

007 cropped

Obviously, the narrowest orifice that the exhaust gasses must pass through on their way out is neither the combined cross-sectional area of the header pipes (approximately 20 cm square) nor the collector’s outlet (also approximately 20 cm square).  It must be then either the round inlets/outlets of the silencer’s pipes or their flattened ends.

Doing a rough calculation, we find that the cross-sectional area of the 1.375” (3.5 cm) diameter silencer’s pipes is 10 cm square, whereas the cross sectional area of the flattened pipe is approximately half that, or about 5 cm square.

What we learn from this exercise is that the HONDA CB550 engine of 544 cc displacement is exhausted through a number of orifices the smallest one of which is toward the end of the exhaust system and has  a cross sectional area of 5 cm square.  That is roughly 1 cm square of tailpipe cross-sectional area per 100 cc of engine displacement.

This is an interesting observation.  I got curious and decided to see if this “rule of thumb” holds for other vehicles.  I measured the tailpipe diameter of three different cars (a BMW, a Chevrolet and a Mercedes ranging in model years from 1986 to 2012), calculated the cross-sectional area and, strangely enough, they had approximately the same ratio of about 100 : 1 for engine displacement vs. tailpipe’s cross sectional area.

It seemed that this 100 cc to 1 cm square “rule” was somewhat universal.  It probably provides reasonably good flow dynamics with minimum back pressure and good scavenging.  And I wanted to maintain it for my bike and it’s new exhaust system.

I bought 4 tulip (trumpet) shaped aftermarket mufflers.   They had a built-in perforated core and a tailpipe that was 1.875” (4.8 cm) in diameter giving a cross-sectional area of 17.8 cm square. The perforated core was big and it seemed to have a very large surface area available to flow.  So, I concentrated on the tailpipe’s diameter and its cross-sectional area.  This observation that it had a a cross-sectional area of 17.8 cm square was shocking:  a single muffler had 3.5 times the cross sectional area of the original muffler.  And I had 4 of them!  This is a huge combined opening with a cross-sectional area of approximately 71.2 cm square seemingly sufficient enough to suit a 7,000 cc engine.  I only had a 544 cc engine, so the cross-sectional area had to be reduced significantly in order to bring it to the 100:1 ratio.

To do that, I bought a perforated sheet of 20 gauge 304 stainless steel.  It had holes that were  5/32” in diameter and were staggered at 0.250” leaving 35% of the total area open to flow.  I cut 8 circles approx. 2” in diameter and connected them together in pairs using M6 stainless steel screws and then made 4 stainless steel brackets to hold each of the 4 pairs to the mufflers.  The result looked like this.



Inserting these into the mufflers’ tailpipes would reduce the surface (cross sectional  area) by 65% through the first disk and then again by an additional 65% thought the second disk thus resulting in the tailpipe’s effective cross-sectional area of 17.8 x 0.35 x 0.35 = 2.2 cm square per muffler. For all 4 mufflers, the combined cross-sectional area that the engine’s exhaust gasses will move through would be roughly 8.8 cm square.  This was still some 70% larger than the 5 cm square or so that I thought I needed and the 100:1 ratio required.  But, I decided that some additional restriction would come from the “torturous path” the gasses had to move through as they exited via the perforated original core the mufflers came with and the disks I installed.

Here is the finished look:



I’m happy to report that the bike runs just fine.  (I did re-jet the carburetors from a 38 slow jet and a 98 main jet to their 40 and 100 counterparts.  For more on that, see the KEIHIN CARBURETORS REBUILD post.)

Here is a list of what I would want to know next time before buying mufflers:










LENGTH OF MOUNTING BRACKET (and/or other mounting details)