# Design Efficient Exhaust System for Project Car

• smoque
I can't seem to find anything specific on the internet, but maybe that's because it's not a common design feature.Since the gases are cooling, it gets slower as they travel thus the reason I am considering a converging pipe, true?What do you think happens when the expanding gas has to enter the reduced cross section at the end of the bend?Remember, the gases are cooling and expanding as they travel.

#### smoque

i wish to design an exhaust system of my project car for minimum flow losses

lets not focus on the exact figures just yet, just wish to confirm whether my design theory would work

the setup would be something like this

exhaust header > 7" length downpipe > 90 degree elbow > 20" straight pipe > muffler > tailpipe

my plan is to let the piping start out large after the header converging into smaller pipe to gradually increase the exhaust velocity to aid exhaust expulsion. plus losses in converging pipes are lesser than diverging pipes, true?

or should i start small and use diverging pipe after the header instead so the maximum velocity of exhaust gas would be near the exhaust header?

which of the above would work better in terms of exhaust expulsion?

i am just wondering is it best if i use one or two sizes larger diameter than the rest of straight pipes for the bend to minimize losses? the piping will be mandrel bent

smoque said:
...

exhaust header > 7" length downpipe > 90 degree elbow > 20" straight pipe > muffler > tailpipe
Any direction change impacts flow. Use a sweeping curve with the largest radius that will fit.

my plan is to let the piping start out large after the header converging into smaller pipe to gradually increase the exhaust velocity to aid exhaust expulsion. plus losses in converging pipes are lesser than diverging pipes, true?
Remember, the gases are cooling and expanding as they travel.

or should i start small and use diverging pipe after the header instead so the maximum velocity of exhaust gas would be near the exhaust header?
Headers with increasing diameter have been tried with various levels of success. The cost in design, testing, and manufacturing difficulty doesn't match the minor benefits. Also, the exhaust velocity and volume is constantly changing with engine rpm and load. The exhaust system design has to be a compromise.

i am just wondering is it best if i use one or two sizes larger diameter than the rest of straight pipes for the bend to minimize losses?
What do you think happens when the expanding gas has to enter the reduced cross section at the end of the bend?

Remember, the gases are cooling and expanding as they travel.

since the gases are cooling, it gets slower as they travel thus the reason I am considering a converging pipe, true?

Headers with increasing diameter have been tried with various levels of success. The cost in design, testing, and manufacturing difficulty doesn't match the minor benefits. Also, the exhaust velocity and volume is constantly changing with engine rpm and load. The exhaust system design has to be a compromise.

the header has same diameter throughout, the diverging pipe i mentioned was after the merge of all primary pipes. something like this:

what is ur opinion?

What do you think happens when the expanding gas has to enter the reduced cross section at the end of the bend?

what if i use a long tapering cone at the end of the bend? the reason for bigger diameter bend is i am trying to get the bend to have a flow equivalent of straight small pipe (i.e. 2.5" 90 degree standard elbow might equal to 1.6" of straight pipe). problem is i don't know what is the good rule of thumb for approximation, perhaps u could enlighten me?

smoque said:
...

I Googled for a reference on tapered headers, and it looks like there are still some companies offering them. There should be some technical references that can better answer your questions. I had a book dealing with racing exhaust system theory and design, but it disappeared long ago, and I can't remember the title.

You can also try calling some of the exhaust system companies and talking to their engineering/technical departments. I'm sure you'll find someone willing to give you some advice.

what if i use a long tapering cone at the end of the bend? the reason for bigger diameter bend is i am trying to get the bend to have a flow equivalent of straight small pipe (i.e. 2.5" 90 degree standard elbow might equal to 1.6" of straight pipe). problem is i don't know what is the good rule of thumb for approximation, perhaps u could enlighten me?
Doesn't matter how long the taper, you're still transitioning to a smaller cross section.

I was taught to avoid sharp bends. The analogy being, imagine a crowd running down a hallway that makes an abrupt turn -- as they approach the turn, the people in front slow down (increase pressure). Now imagine the hallway makes a sweeping turn -- the greater the radius, the easier they transition. If you are stuck with making a tight 90° turn, there may not be a simple solution.

I have a vague memory of experiments with dimples and/or bumps on the surfaces of a bend to influence the boundary layer and improve flow characteristics. You might want to look into that.

I wish I could be more help. It's been quite some time since I worked with this stuff, so I'm a bit fuzzy on the details.

what if a pretty tight 90 degree bend is required? due to space constraint?

will using larger diameter 90 degree bend helps to smooth out flow? i.e. 1.5" straight pipe > 2" diameter 90 degree elbow > 1.75" straight pipe > 1.6" straight pipe > 1.5" pipe

instead of 1.5" > 1.5" 90 degree elbow > 1.5" pipe

Are you familiar with Fanno and Rayleigh flow theory?

no I am not aware of that, how is the theory relates to what i need?

here's the link to raleigh flow in wikipedia
http://en.wikipedia.org/wiki/Rayleigh_flow
in first paragraph click on Fanno flow
it will take you to that section
have the 'exhaust system' be as equal as
possible to the sum of the individual header tubes
you will avoid 'reversion' and 'constriction'
it involves calculating area of a circle and addition