Intercooler Piping Routing Under the Car- Pic

[ktm]

IAT won't be a problem as I am going to run the Wolf EMS. I am running a T03/04E 0.63 A/R, 50 trim turbo. I'll be boosting from 15 to 20 psi.

 

Welding the connections is something I may do in the future, but in all honesty there are only two connections that would be eliminated by welding. I still need the TB and both IC connections, as well as the turbo connection.

 

Thanks all for the compliments.

[WizardBlack]

The air is moving at over 300 ft/sec.....an extra couple feet is nothing. The slight increase in volume is the other issue, but that would only impact the turbo lag. Again, given the velocity, the flowrate would be such that the incremental increase in volume would have a minimal impact.

 

As for the bends, there are no more bends in this setup than there are in a convential behind the fan setup - 4 bends for the inlet, 3 bends for the outlet. As a matter of fact, my outlet bends are very smooth. One 90 degree and two 45 degree. The bend coming up from under the engine is a 70 degree bend.

 

300 ft/sec = 18000 ft/min

With a cross sectional area of pi*2" for two inch piping, that's 6.28 in^2 or .04361 ft^2.

So, for your estimated average air velocity, you would have a flow rate of 785 cfm.

I think the piping velocity is probably more about half of that; or 150 feet/sec.

 

Aside from which, I don't care what Corky Bell or anyone else says about what pipe diameter is efficient (even though he puts out a pretty good book; I have read it). Does he define efficient? Will 2" work for 785 cfm engine, sure. Would I run 2" IC piping on an engine flowing that kinda air? Hellz no.

 

BTW, the extra length is actually pretty irrelevant as far as volume is concerned. Just keep the bends minimal and the heat low.

 

Looks good, BTW.

[ktm]

Two inch piping? Nope, but I am running 2.5 inches. The turbo outlet is 2-inches; I am running a 2 to 2.5 inch reducer right off the turbo outlet and 2.5 inches all the way to the TB. At the TB I have a 2.5 to 2.75 inch reducer (my TB measured 2.75 ID).

 

(2.5"/12 inches/foot /2 (for radius))^2 * 3.14159 = 0.0341 sq. ft.

 

Assuming 600 cfm (400 hp - again, referencing CB's book), that yields 600 ft3/0.0341 ft2 / 60 secs/min. = 293 ft/sec.

[Pyro]

Yes, the turbo output OD is 2 inches but the wall is about 1/4 inch thick. So, all that air flowing out of a 1.5 inch hole. Makes me wonder how big a pipe is really required.

[Pop N Wood]

Not sure I want to get back into this but you guys do realize that air is compressible?

[Clifton]

The extra length doesn't add anything noticable in lag. I have probably another foot of tubing on mine compared to KTM's, 2.5 in, 2.75 out, 4" IC and that is up from the 2" tube and a shorter run with a different 2" IC that I originally had on it. It wasn't noticable at all.

[WizardBlack]

Not sure I want to get back into this but you guys do realize that air is compressible?

 

I would hope anyone posting in the TURBO section would know that.

[Tony D]

Lag is overblown. My Datalogging with the Megasquirt shows lag to be imperceptibleonce boost threshold is attained. My WOT line traces almost like a digital switch to full boost from vacuum. Most 'lag' is really from improper driving below boost threshold. I think the setup has deminimus effects from the piping route chosen. For the available horsepower goals, I would be suprised that the effects will be noticable at all. Given the numbers KTM calculates, in the real world they mean 'nothing'...this is a theoretical argument at this point, and I will lay money that in the real world of operation, the setup will function flawlessly. Regardless of what theoretical improvement might be available.

[Pop N Wood]

I would hope anyone posting in the TURBO section would know that.

 

So then what does taking 600 cfm and dividing by the cross sectional area of the pipe yield?

[Cruez]

I don't think that lag would be much more than a conventional setup... maybe at least not seat of the pants noticeable... They have setups now that are running the turbos at the rear of the car and they mentioned that lag was not a problem... You might have some pressure drop, but its all in what you want the car to do and look like.. IMO

[24OZ]

Looks very nice indeed. I will have to do mine soon and it's good to get some ideas. Thanks for posting.

 

One question, from the turbo did you use a straight silicone expander and then a 90 degree pipe or a 90 degree silicone expander and then a straight pipe? Just wondered if it makes a difference?

[ktm]

Straight reducer and then a 90 degree pipe. A 90 degree reducer would hit the steering shaft and would also not give you the clearance you need to run under the oil pan.

[jgkurz]

That's a seriously clean looking setup. I run my IC out pipe parallel to the radiator which has worked well. My problem is the cooling fan clearance. I'd like to run a clutch fan instead of an electric fan. They're way more reliable in my opinion.

[e_racer1999]

^agreed. clean setup, and looks like it will work just fine

[WizardBlack]

So then what does taking 600 cfm and dividing by the cross sectional area of the pipe yield?

Average airspeed in the pipe; 300 ft/sec. I wasn't correlating 600 cfm to 400 hp when I typed that post for some reason, but it comes out to 300 for 400 hp through 2.5" pipe. However, as I think of it now, wouldn't that be 600 cubic feet of standardized pressure air? If it's compressed it would actually be a lower cfm rating, albeit at boost? I dunno; it's been a long day. heh.

[Pop N Wood]

Average airspeed in the pipe; 300 ft/sec. I wasn't correlating 600 cfm to 400 hp when I typed that post for some reason, but it comes out to 300 for 400 hp through 2.5" pipe. However, as I think of it now, wouldn't that be 600 cubic feet of standardized pressure air? If it's compressed it would actually be a lower cfm rating, albeit at boost? I dunno; it's been a long day. heh.

 

This was the point I was after. You cannot compute flow velocity without knowing the pressure and temperature of the air. Bernoulli's equation will tell you how to get the correct number. It will also tell you the flow velocity varies continuously in the intake track because pressure and temperature are not the same at any point in the intake track.

 

So in that sense KTM's calculations don't tell you anything about what an extra 100+ cubic inches of intake piping will do to the motor.

 

What was confusing me is why everyone is so keyed into flow velocity. I am not sure what you think that is telling you.

 

As for lag, if the only way you use the throttle is to mash it to the floor at the start of the run and to let off again at the end of a quarter mile, then you guys are right. Once the turbo spools up and the motor is at full power than that extra length of pipe probably doesn't matter.

 

But if you actually try to use the throttle and engine power to negotiate the car through a turn, then I simply cannot believe that increasing the volume of the intake plumbing by an amount almost equal to the engine displacement won't be noticeable.

 

Think of the intake plumbing as a storage reservoir on an air compressor. The larger it is, the longer it is going to take to pressurize, the more stored energy it will contain, and thus the longer it is going to take to depressurize. None of those things will contribute positively to throttle response.

[ktm]

So in that sense KTM's calculations don't tell you anything about what an extra 100+ cubic inches of intake piping will do to the motor.

 

 

All turbo charged cars with intercoolers have intercooler piping. Comparing a turbo charged car with an intercooler to an N/A car is an apples to orange comparison.

 

As for lag, if the only way you use the throttle is to mash it to the floor at the start of the run and to let off again at the end of a quarter mile, then you guys are right. Once the turbo spools up and the motor is at full power than that extra length of pipe probably doesn't matter.

 

*snip*

 

Think of the intake plumbing as a storage reservoir on an air compressor. The larger it is, the longer it is going to take to pressurize, the more stored energy it will contain, and thus the longer it is going to take to depressurize. None of those things will contribute positively to throttle response.

 

....and the time to fill that reservoir is a function of the flow rate. A flowrate of 600 cfm is equivalent to 1,036,800 cubic inches (600 ft3 x 1728 in3/ft3) a minute.....

[Clifton]

But if you actually try to use the throttle and engine power to negotiate the car through a turn, then I simply cannot believe that increasing the volume of the intake plumbing by an amount almost equal to the engine displacement won't be noticeable.

 

I promise, it's not noticable. I have no problems autoXing, going from 26 psi to closed throttle, to part throttle, back to full boost at the limit of lateral traction. If the throttle plate was at the turbo inlet instead of the manifold I think it would be completely different. I can say that when I ran a afm just before the turbo, it wasn't quite as smooth with throttle transitions with the length between the throttle and afm being so far. With a map sensor at the plenum, it's perfect.

[Pop N Wood]

I promise, it's not noticable. I have no problems autoXing, going from 26 psi to closed throttle, to part throttle, back to full boost at the limit of lateral traction. If the throttle plate was at the turbo inlet instead of the manifold I think it would be completely different. I can say that when I ran a afm just before the turbo, it wasn't quite as smooth with throttle transitions with the length between the throttle and afm being so far. With a map sensor at the plenum, it's perfect.

 

Well you sure have a drop dead gorgeous car, whereas I have never built a turbo motor. So a good point of reference for me. 26 psi seems like a lot of boost, too.

 

Do you any type of pop off valve to relieve intake pressure on quick down throttle?

[Pop N Wood]

All turbo charged cars with intercoolers have intercooler piping. Comparing a turbo charged car with an intercooler to an N/A car is an apples to orange comparison.

 

The only apples I care about are going around a corner as fast as I can. Your statement makes no sense.

 

Besides, just to be difficult, a turbo charged car with an air to water intercooler doesn't have to have piping. That is the reason some people build their cars that way.

 

....and the time to fill that reservoir is a function of the flow rate. A flowrate of 600 cfm is equivalent to 1,036,800 cubic inches (600 ft3 x 1728 in3/ft3) a minute.....

 

Well not quite correct. Filling the reservoir is function of mass flow rate. And once you start thinking mass flow rate, then you need to ask yourself what you think you mean by CFM.

 

Did you look up Bernoulli's equation yet?