Roman

Oh yeah just remembered. When you put it back on, there are some valley areas that need sealant, and some that if you fill them up your motor starves of oil haha. So take VERY CAREFUL note of which areas under the big cam cap have sealant on them.

You need to take the cam gears off. So you need a large hex head of the right size which undoes the front cap, and then there's another large hex under it which takes the gears off Then there are bolts for that front cover under there, then it all comes off. Before you do anything further, put those pulleys back together ASAP or they'll explode across the room and you'll never get them back together

Not mufflers as such, but check out this vid... very interesting!

Haha, no graphs but give it time... It seems regardless of whether manifold is aftermarket or factory, the total length to the 2>1 or 4>2>1 seems to be about the same in each case. Which means the extractors finish somewhere under the passengers feet. So probably impractical to make a 4:1 thats any good. Could probably keep the existing 2>1 section and just make a different 4>2 and see how that goes. To be honest though I've got no idea how you can quantify the change in how an exhaust manifold does its thing. Would the fuel requirements change? Or is it just one of those things you cant tell without a dyno.

I've been thinking about trying to make a different exhaust manifold, basically to try and have a longer Straight section on the runners before the first bend. There's heaps of space for it. Currently 1 runner has the bend immediately after the flange and the other is maybe 100mm or more away from the flange. Also, my exhaust ports are larger than standard and I had to die grind the hell out of the manifold to be the right size. But it steps back down to smaller diameter which probably isnt great. So it would be interesting to see what happens going up a pipe size. For interests sake thought I'd draw up some bends in solidworks and see how close I could get to drawing up an equal length exhaust manifold. In my mind it's the same process that you'd do on the bench tacking things together but a lot quicker to readjust angles. Still took ages, would take a zillion times longer IRL! I thought I might end up drawing a model that needed a zillion cuts and welds and bends to work, but it looks as though with the magic of CAD I could probably bang something together from off the shelf mandrel bends, only 3-4 parts per runner to get equal length within a few mm. Based on the test model, wtih some accurate cutting and coming up with a way to index the angles between parts from the model. Only 3 parts per runner would be required: (excuse the wonky angle blue pipe, solidworks crashed before I saved it and CBF redoing it) I reckon it looks feasible enough to spend a bit more time modelling engine bay available space, exact port spacings etc and mock something up. Anyone tried making a manifold from cad drawings before? Might pull the manifold off and measure some stuff in engine bay next time there's a rainy few days.

Hah. the bread bin rules.

Has these weird lines through it... Cant see it from the inside, but it's annoying. concerned that it's delaminating or something like that? Good idea to get screen replaced, or not to worry.

What do you think will happen with larger dia pipes? I'm guessing it would lower the tuned length rpm of the manifold, all other things being equal Or is it more about just hoofing the air out a bit better.

His 4age extractors wont fit the 3SGE that he's mounting in the car now, so he's remaking them.

If you plumb it back into pre-throttle body somewhere, when it makes boost the pressure on both side is balanced so no leak. Depends on the ISCV unit I guess but not plumbing back to pipe and you'll have air blowing back out through it.

Your sample size is 1, so it's the worst option you've tried as well as the best I was going to get a 14point7 unit but didnt want to wait on shipping so got an LC2 locally same day instead.

There are two types of Bosch sensors commonly used in most of the wideband kits (LC2 etc) there's the 4.2 and 4.9 The earlier one (4.2) is prone to shitting itself after not too long (mine lasted... 6 months? 9 months?) where as the 4.9 is actually used in a lot of OEM cars now because they figured out how to make them last long enough to do so. After my 4.2 shit itself I upgraded so now it'll hopefully last a lot longer. But just any FYI / the poor man pays twice / etc Not worth saving $30 or something and ending up with a sensor that doesnt last too long.

Modelled my fuel lines, rail, hot bolts, fuel flow rates from pump etc etc to see what happens. Looks as though just hot bolts holding the rail on are enough to send fuel temps up and up as the fuel circulates. However adding a cooler with a little airflow through it and problem solved. Soooooooo..... I'll do that at same time I fit an adjustable FPR to get the pressure up a bit higher. Once I've got the fuel temperature and fuel pressure sensors fitted (as well as an adjustable FPR for higher pressure) I can switch over to a different fuel equation type in the ECU that should help keep it more accurate with changing conditions. Not that it's awful currently, but something interesting to learn about.

My calcs show that if the pressure waves are going at speed of sound, my runner length must be 600mm, which definitely isnt the case. I'd say it's maybe 350mm total length or something like that. But the shape of fuel graph still looks as though its happening I think. So my thoughts... The pressure waves dont actually travel at the speed of sound, maybe half that speed.

Applying my shoddy maths to the above on the 75mm graph. You've got a peak at 3200rpm and a peak at ~4700rpm, so ratio of 0.68 So looking at my numbers means: Must be the 2nd and 3rd harmonics making these peaks So doing some wizard magic in my thingy, your other harmonic pulses happening at: Next one coming up to 9500rpm Also notice where it says that theoretically one of the "troughs" is at 6300rpm which is what it shows on your fuel map too. If you went 10mm longer it goes like this 20mm longer (to 95mm) like this 50mm longer (to current) like this (Chucking air back out at 6500rpm....) 95mm sounds good by the maths too, because it pushes that trough below 5500rpm

Yeah keeps gaining fuel but I'm already pushing my luck with 8300rpm haha so not keen to test any higher. I'll PM you in a bit and send you some stuff. So, on the topic of peaks and troughs, have found something really interesting. If you do a manual calculation of Helmholtz frequencies, as in, calculate the length of time it takes a sound wave to go from the back of the valve, reflect at end of the runner and then get to the valve again. You end up with an RPM that the first reflected wave hits. Then the wave travels back up and reaches the valve a 2nd time, (weaker pulse) then 3rd, then 4th. (weaker pulse each time) What's interesting about this though is that If you change the runner length, or the speed of sound, the ratio of time between the 1st, 2nd, 3rd, 4th etc peaks stays exactly the same. Soooo this is useful, because if you look at the rpm distance between the peaks on your graph, you can tell which harmonic wave it is that's causing the peaks and troughs. And then you can extrapolate out to find the others. So in my instance, a peak at 4000rpm and a peak at 5500rpm. 4000/5500 = 0.72 so it must be the 3rd wave reflection happening at this RPM? So if I play with my calculator, to see if I can find what runner length it believes is necessary to get a 3rd wave reflection occuring from 4000rpm to 5500rpm 59.8cm effective length from valve to runner end gives a peak that's only 1300rpm wide, not 1500. No matter what I do, changing runner length or speed of sound I cant get a 1500rpm wide peak to peak starting at 4000rpm. (note how the standard intake manifold has a 2nd harmonic peaking at 8100rpm, good times had by all) So then I thought... Maybe its because you dont want the wave to return 720 degrees later (as the valve is already shut!) but something earlier like 650 degrees. But this just makes the same difference as changing the runner length, so wouldnt widen the peaks. So then the next thought... Not only are there "good" reflected waves, but there are also ones where the reflected pulse pulls the fuel back out of the motor instead. Happening at twice the frequency. So if I make an RPM plot of where both the peaks and troughs theoretically occur, and then subtract the peak strength from the trough strength to get an "effective" wave strength. The effective peak to peak becomes slightly longer or shorter because of some instances where waves cancel each other out. The lines on this graph below should probably be sine waves instead of pointy, but I dropped out of 6th form maths so dont know how to make a sine wave in excel The blue line is the "good" pulse. Red line is the "bad" pulse and the green line is the "effective" pulse, AKA what I'd perhaps expect to see affecting the fuel map My graph is shitty and I've made some heinous assumptions about wave strength etc. But you get the idea. Check out that peak and trough that both happen at about 8100rpm. Maybe this is why intake manifolds are tuned for 3rd harmonic instead of 2nd, the 2nd cancels itself out.

I dont really understand the units or maths of what's written below, but came across it somewhere and thought I'd jot it down and come back to it if I manage to get my head around it.

You're right. I frigging love it. Post all of the things!

I've got these peaks and troughs in my fuel map. Figured it would relate to intake manifold runner length. I was given some other fuel maps from various beams engines, found one of them was interesting. (Has ITBs) These are the fuel maps at WOT for both motors, mine on the left: So the peaks and troughs have moved to the left on the ITB engine, which is strange because it would suggest the tuned length of the runners is longer than mine. I asked the guy I got the maps from, and he confirmed it had crazy long runners. It looks like the fuel tapers off towards redline because it's heading into the third resonant "trough". He said some time after this they shortened the runners back down and gained power everywhere. It would be interesting to see the fuel map of a motor with ITBs with short runners and see how the fuel map moves around.

Interesting! Will love to hear how you get on.

Hmmmm okay interesting about the heat difference thing, makes sense I guess. Yeah one of the big reasons for a returnless system is for emissions, when you sent hot fuel back to the tank it creates a lot of vapour which then either comes out the gas cap when you open it, or needs to be dealt with by burning it through the engine which I think creates less than ideal emissions thingys as well.

In relation to above, Common Rail diesel cars generally have a fuel cooler in place, buuuuutttt its always placed on the return rail back to the tank, rather than cooling the fuel coming to the rail. I think this is more credibility towards the idea that it's generally the rail that heats up the fuel. This way you're cooling it just after it's been heated, so its not hot back at the tank and second time around. Also minimising complexity of the high pressure side. Fascinating topic! ... ...

That's some impressive axle twisting going on! Hope it's an easy fix.

The effective width is still 100mm in any case, because the fins in the radiator still run perpendicular rather than sloped. There's still the same surface area of fins cooling the same volume of water. I'd say it's more about fitting a bigger radiator while keeping that nice low bonnet line that RX7s have. I've seen some racecars that have the radiator mounted down flat, with ductnig to push air through. I'd say any angle is fine so long as you're getting air through it, and you can bleed the coolant.