Upcoming Events for the Next 14 Day(s)
How to match a manifold.
Its common to find two sections of a split intake manifold and/or the manifold to head ports do not match particularily well. Sometimes this can be by a very minor amount, sometimes by a large amount which will rob power by impacting flow through the manifold and into the ports in the head.
This miss match reduces flow as it creates turbulance within the manifold, a large to smaller change in the direction of flow is by far the worst form of miss match but a small to larger change also won't do it any good. In an idea world you want both sections to match as closely as possible.
At best matching the manifold and head will release a couple of bhp, however it can greatly help cylinder filling at low engine speed.
So as I was matching the manifold on our race car today I thought I'd do a quick how to. On our Clio the match between the lower manifold section and the head was pretty good, good enough not to worry about anyway ;-) The port shape between the two manifold sections on the lower manifold was also good and to be honest it appears the lower section is a much nicer casting than the upper...... which is rubbish, albeit not so rubbish that it is larger than the lower casting.
- Kit to remove the manifold
- Engineers Blue
- A Dremel or similar with a couple of decent grinding stones
- Big Wipes or similar grease removing wipes
- Parts washer or a dish washer and an understanding other half
Heres the upper and lower manifolds together, you can see what a poor port shape the upper has and how it differs to the lower. As the lower is pretty good I'll be matching the upper against the lower. There is sufficent material to do this without removing too much wall thickness but you should always check this!
Just to show how much they differ here's the lower section.
And here is the upper section. Notice the massive difference in port size!!
Certainly not ideal! So its time to sort it! Time to turn your skin, clothes and everything else you touch blue
You'll need to smear a very thin but even dark blue layer onto the section of the manifold which is the 'correct' or larger size. If all the ports on one section are not larger i.e two on the upper are larger but two it are smaller then you will need to blue up accordingly with the larger ports blued.
Once blued bolt the manifold sections back together with enough torque to squeeze the blue. It won't spread unless you toque the fasteners up to silly amounts. All you want to do is make sure the entire surface of the manifold sections are in contact.
Then when you split the manifold sections you'll have a perfect imprint of the larger ports around the smaller ports. You can see how bad the match is!!
Using the blue as a guide you need to open up the ports to match. It is important you avoid creating a step or disturbance within the manifold/runner so if you need to remove material also smooth the area behind it to maintain flow, it's also a good idea to remove any casting flash or nasty bits of casting whilst you are in there with the dremel.
Remove small amounts of material at a time, you do not want to over do it and a dremel with some force behind it will go through the super soft alloy that Renault case the upper out of in no time!
First bit of dremeling complete to remove the majority of material required.
Heres a close up view showing a nasty casting line in the runner.
Once the initial removal of material has been done reblue and recheck the manifold to get the port shapes correct.
Still a bit to do here! You'll need to reblue and recheck several times. Don't be tempted to rush it or you'll cock it up ;-)
Bit more work and it's looking much better. Just need to chase that casting line back in that runner now and take a little bit of material off the left hand wall.
About done, note the much nicer port shape than the original.
Just need to reblue for a final time and check the match.
Now that is lookin a lot more like it! You can clearly see how far I've had to chase back to get a decent shape. Unfortunately it wasn't possibe to get all of the casting lines out as doing so would have meant expanding the runner diameter beyond the port diameter and in the grand scheme of things there would be no gain to be had.
Nearly got it all out of this one though, just a shame it runs up so close to the port. Can't nibble any more out of that without risking damaging the port shape and therefore match. If you really wanted to remove this you would reshape the port accordingly and blue it up against the lower section, reprofiling the lower to suit.
Easy as pie see ;-) If you are doing it take your time and go slow. Do it bit by bit and you won't go wrong - even if you do get a few spots wrong the chances of it being worse than the standard manifold are minimal!!! LOL
It's a lot easier if you have a CNC mill and a port shape design as you simply program that to buzz around both manifold sections and you get a pefect match! However as it takes a lot of time to write the CNC program/do the drawing its not really practical for one offs.
Just a word of warning as the material is so soft it will burr up around the ports and its very easy to catch the flange, as such you need to budget the time to have the manifold skimmed afterwards to remove the burs and put a flat surface back on it before refitting.
Last edited by Icarus; 04-04-2008 at 19:40:54.
woahhh?? is it me or did i read that 4 times while scrolling down once??
Just had a look, think it's alright?
Top effort there mate!
But why don't they make the cast pattern moulds the same at the manufacturing stage, the port patterns at least?
Then just a quick grind and polish surely?
They measured about +3mm difference in tolerance there (between top and bottom ports)
Tell me about it mate. The casting is done on the basis that x amount will be removed from the fact of the flange. Tolerance for this is always pretty big though hence where most of the mismatch comes from.
Also the moulds will have a life so if the upper is at 90% and the lower is at 10% of life you'll get a difference from that as well.
OEM's have got much better at it in the past few years and CNC finishing on heads etc. is pretty common now which makes up for poor casting quality, indeed can be cheaper to have ****e cast quality and CNC finishing than a good cast quality in the first place.
The lower manifold is fine, just the upper that is rubbish. Not sure if this is the same on all of them..... or just ours LOL
Strange thing is though, on the pic' where you first separated the manifold after blue-ing, the difference around the port edges is almost perfectly uniform?
Like the ports are almost exactly the same shape but one set is simply physically bigger????
Just seems the cast pattern of one, is massively different to the other???
I'd guess at wear mate but anything I've done with casting in the past has been a thousand units max hence you never really get into issues with pattern/mould errosion.
I wonder if its to do with the amount the flange is machined. The runners increase in diameter as they run down so if it is machined more/less than blueprint you'd end up with a difference in diameter.
Might have to call a few peeps now as its perked my interest up as to why its so ****e LOL 172 and 182 are the same by all accounts.
Would explain it definitely. Didn't realise the chamber was larger the further in/down you go. Good spot mate.
Originally Posted by Icarus
Nice write up there, Icy. Would be a nice little CFD study to model the two haves of the manifold before and after matching and see the difference in flow characteristics.
p.s Doesn't mean i'm offering
Sorry I meant decrease but the effect is the same if you skim them. They get smaller to keep the gas speed up at low engine speeds. Definately going to look into it further. Purely out of interest though as it doesn't really effect the low volume stuff I do LOL
Originally Posted by Hair Bear
Would assume its purely cost based to be honest and the tolerances are so big to avoid binning/recycling so many manifolds.
Users Browsing this Thread
There are currently 1 users browsing this thread. (0 members and 1 guests)