SDI Intake Manifold Installation

[fspGTD]

No, it wasn't there originally, but the goop was applied just an experiment to trace IF there was a boost leak through the seams of the moulded manifold halves.  There doesn't seem to be any... so I believe it's holding up under boost.  I meet with Gerry again this weekend to see if we can put this one to bed.

In the final analysis, was it worth the $327 US?  HP per dollar, no.  I'd sooner install an FMIC or a modded pump, or bigger injectors, or an after market cam, or custom chiptuning.  But I knew from the very outset that the HP gains from this manifold in and of itself would be negligable.

Jake, if you're looking for real gains in your ride, class rules permitting, I'd sooner adapt a VNT-turbo into your car and get a Giles pump.  Can you put modded / bigger injectors into an IDI motor like you can on the TDIs?

Um, really in a competitive race classing situation, the "real" gains are no longer available as they've already been done, assuming the rules are anywhere near fair!  I am already pretty competitive at this point, and am just focusing on fine-tuning.  IE: any honest-to-goodness 3% power gains that I can get without too much work, I don't want to leave on the table.  So, I am focusing on what others might not consider to be real gains, although for me something like a 3-6% honest to goodness increase in peak horsepower is VERY noticeable and helpful on the track.  If I can combine together a few small gains... they all work together and what do you know it, the overall system is getting FASTER.

My autocross racing class is actually quite restrictive, although that can be viewed as a good thing, for helping to somewhat keep a cap on the budget.

First and formost - this surprises most turbo-diesel folks: It doesn't allow any changes to wastegate or turbocharger.  Engine internals must remain unmodified (only blueprinting within the factory specifications are allowed) and camshaft must also remain stock.  The exhaust system is free, complete fuel injection system is free, intercooling is free, and also the manifolds (both exhaust and intake) are free.  I have modified fairly extensively (of course not completely... but there are diminishing returns) all of the above with the exception of the manifolds, which I have only port-matched to the head (this is allowed up to 1" depth into the head), and I have ceramic-coated my exhaust manifold to maintain heat inside and through the turbine section to reduce turbo lag.  I have an audi5000 intercooler that is highly efficient, and does not restrict boost pressure measurably, 2.5" exhaust system in combination of stainless and lightweight aluminum with a custom downpipe from the turbine outlet back and no muffler, 1.9lTD 2-stage injectors and have done some dyno testing with change the governor mechanism on my fuel injection pump, but the pump remains the stock plunger 9mm bore and stroke.  It is at least close to being rich enough though!  And it is costly going to a bigger or better pump setup (believe me, I am considering that.)  About the only thing I haven't done besides extensive blueprinting work (who knows I may be going down that route soon, as I may have to build another motor to replace my old damaged one) is optimize the intake and exhaust manifolds.  Since changing the exhaust manifold is more difficult and expensive than the intake, I was leaning towards keeping the stock exhaust manifold, and along with that not change the turbo mounting location, exhaust system from turbine back, turbo oil return, and air intake system.  So I have high on my list of performance optimizations upcoming, replacing the factory intake manifold with a custom long-runner unit designed to optimize the usable power curve.

Sure, more max hp would be great if that were possible, but my car I think is powerful enough in my class, and it might be unrealistic to expect large power gains from a long-runner intake (something like a 3% power gain might be possible and would be VERY welcome, of course!)  What I would hope to get though primarily with the long-runner intake manifold, is a broadening of the usable torque curve down lower into the RPM range, basically to make it so I can keep it in 3rd gear through slower turns without it lugging and making me shift to 2nd.  Currently when the RPMs drop down to about 2500-3000rpm at speed in 3rd gear, the engine bogs as the turbo is not spooled up and engine needs to accelerate more before the turbo is working to a point where it comes "on-boost".  I'm thinking a well-engineered custom long-runner intake manifold could help the turbo start working in this area where now the motor bogs, and thereby extend my usable RPM curve down lower into the RPM range.  Sure, any benefits in HP up higher would be a nice side benefit, but that's not my primary goal.  A 3% power gain on the side would be nice though and I will certainly take that if I can get it!  OTOH, a significant restriction in mid or high-RPM airflow, or much power LOSS in that area, would be unacceptable.  An objective with the long-runner intake manifold project would be to not see a significnat reduction in power at any are in the powerband.  (down at idle, I don't care... up high in RPMs above the revv limiter, I don't care right now, although I may be experimenting with revving the engine higher as well, but that is another project.)

Maybe I should also point out that if my power curve were beefier down at a lower RPM, at the autocross school where my car broke, I may have been able to have left it in 3rd gear in the section that I had to run it in 2nd where I mis-shifted into 1st instead of 3rd, over-revving my motor and bending the valves.  Yeah... you see it was the intake manifold that killed my motor!!!  (OK, just kidding... but hopefully that does help to illustrate the need I have for broadening my power curve down lower...  reduce or remove the need to shift!  having to shift a lot in the middle of an autocross course is bad!  There are enough other things for me, the driver to think about... course recall and looking ahead, where does the braking point start (can't left foot brake if you need to downshift at the same time!), where to position the car and turn in, what line to take, are all cones where they should be or should I stop if one's out of place from where I remember it should be on the coursewalk and point it out so I can get that re-run, etc.  if I can remove the worry about "can I get through this corner in 3rd or am I going to have to downshift into 2nd to avoid a bog", etc... it's a good thing!  Anyway, you get the picture.)

[farkman]

TDIMeister

Have you noticed any measurable increase in fuel economy with the manifold?  

fspGTD

Regarding your stock exhaust manifold, have you considered extrude honing? I don't know how much it would be, but at least you don't have to change the location of the turbo.

Peter

[TDIMeister]

Jake, IMHO you might consider looking to adapt a manifold from a non-cross-flow 2.0 gasser.  It also has long runners (albeit no where as long as the SDI ones...  much easier to find and likely much, much cheaper.

The SDI manifold will not stay in my car for long... I have visions of grandeur to build a complete motor incorporating this manifold and a VNT-20 turbo I just bought.  This, and a RedRotors-modified pump, big honking injectors and custom programming will make this a Monster TDI!

Since intercooling and fuel injection mods are unlimited, exploit this to the fullest possible extent!  Plumb a dual-IC setup ora massive unit that covers the entire front-end or something!  What are the rules about changing turbos?  IIRC is must be from a unit that was incorporated in the OEM engine, although you are able to go through the parts bin in case there were different, better units from different years?  The old T3 turbos are relatively unresponsive, inefficient units.  A GT-15 or KKK K03 would give a nice improvement in spool-up time and will have greater efficiencies.  Both were used in later 1.9 IDIs.

[TDIMeister]

TDIMeister

Have you noticed any measurable increase in fuel economy with the manifold?  

I'm right at about my average.  My driving patterns are relatively consistent with about 75% of my kms put on the highway, and the rest on a short 9 km commute to/from work.  I'm getting mid-5s L/100km.  While I don't abuse my car, I don't exactly drive like a granny, either.

So, short answer is: No noticeable difference with the manifold.

[fspGTD]

Jake, IMHO you might consider looking to adapt a manifold from a non-cross-flow 2.0 gasser.  It also has long runners (albeit no where as long as the SDI ones...  much easier to find and likely much, much cheaper.

Let's see... you mean like from an Audi 2.0l reverse-flow block?  Like the other gasser reverse-flow blocks, my turbo sits too high and would interfere with it's layout.  Now I would definitely try that if I were to change my exhaust manifold to relocate the turbo lower and out of the way, but that's a bigger project... re-doing the turbine outlet exhaust system, relocating the turbo, changing the oil return line, re-doing the air inlet tract going to the compressor inlet, not to mention boost tubing coming off the compressor outlet...

The SDI manifold will not stay in my car for long... I have visions of grandeur to build a complete motor incorporating this manifold and a VNT-20 turbo I just bought.  This, and a RedRotors-modified pump, big honking injectors and custom programming will make this a Monster TDI!

Since intercooling and fuel injection mods are unlimited, exploit this to the fullest possible extent!  Plumb a dual-IC setup ora massive unit that covers the entire front-end or something!  What are the rules about changing turbos?  IIRC is must be from a unit that was incorporated in the OEM engine, although you are able to go through the parts bin in case there were different, better units from different years?

With my current setup, intercooler efficiency of the single audi5k intercooler at stock peak boost pressure are very efficient, so little would be gained by adding another intercooler, and there would be the negative of added weight and volume in the intake tract that could actually decrease overall performance.  Also, although intercooler are unrestricted, you can't just go cutting holes, etc and do unauthorized modification to get them to fit.  You've got to carefully place them.  Anyway based on test results of the intercooler outlet air temp as well as pressure loss through the intercooler, I think I am already near-optimal on the intercooling situation.

FYI: Here is the section of the rulebook that has most of the items that cover street prepared class:
http://mouton.best.vwh.net/sccasolo/Rules/sp.html

I can only update/backdate '75-'84 Rabbits and Jettas.  That leaves as the only turbo diesel engine option, the '83-'84 engine code "CY" turbo-diesel 1.6l solid-lifter, for which there were 2 turbos available: a KKK and a Garrett, both of which came with the same peak boost pressure specs (9-10psi).

There are a few other interesting options I've considered, but they are not what most folks do to their street cars.  Water injection is permitted (but no alcohol.)  Also, CNG-fumigation is permitted.  Like I said before, my car is pretty fast as is... I'm not in need of a huge power gain.  Of course, if there is something big I've left on the table, I'd like to find what that is, but trust me I've given this a lot of thought and you can rule out upping the boost pressure (beyond maybe blueprinting the stock turbo to the upper limit of the specification, probably not more than 10psi).

Oh yeah, I do think there is a lot of potential to be had in the fuel injection.  And this is an area I have been and will continue to focus a lot of research time on.  I have seen HUGE power increases in this area already.  (One of the main power boosters that allows me to kick out around approx 120hp on stock boost pressure out of a motor that originally had 68hp!  Not to mention how much higher the power curve extends that the stock motor (whereas the stock governor petered out starting at around 4500rpm, I've gotten mine to pull with full power up to about 5500+rpm!)

Believe me, it's very good that I stay in this class.  Were I to mod my turbo, I'd go into prepared class, where lightened cars trailered-only need apply, with lexan windows, stripped interiors, double a-arm conversions, fully-internally built motors, etc.  Or street mod class, not much better.  Street prepared is where I can hang with the 1.8l "JH" sciroccos, Rabbits and GTIs, etc.  It's a good competitive class where VW drivers have taken national championships.  Frankly I feel in current state of tune, my 1.6lTD was at least as competitive as a prepped to the limit 1.8l"JH" motor, so anything I can do to make it even better will only further the advantage!  On the disadvantages of my setup, a 2-dr version would be approx 55lbs lighter than my 4-dr version.  And a scirocco although not lighter than a Rabbit (actually a little heavier) may be more competitive yet, likely due to a lower center of gravity.  There is a rule change proposal in the works by that way that would combine 8v sciroccos with the Rabbits/Jettas onto the same line... which would mean that an '83-'84 1.6lTD swapped into a scirocco would be legal!

The old T3 turbos are relatively unresponsive, inefficient units.  A GT-15 or KKK K03 would give a nice improvement in spool-up time and will have greater efficiencies.  Both were used in later 1.9 IDIs.

I'm not really going to argue with you there, except that I do think they are quite efficient matched to a 1.6lTD, but only at high engine RPMs.  And I'm sure with modern ball-bearing and/or variable vane technology, one could come up with one that's even more efficient.  I just can't on my car, because there's no way that would survive a protest at nationals.  And I would likely be protested for such a modification, were I to win at nationals in my diesel.  I'm the only one who's raced a diesel at all at SCCA nationals!  There are over 1000 racers competing in nationals each year.  If I were to win my class in a diesel, you better believe I'd have people crawling all over it looking for protest-fodder.  It's also going to make some major waves in the racing community, which I would love to see happen.    Just to get back at everyone who told me that it was impossible for a diesel to be nationally competitive.  

[caddy]

do you think if that intake could resist at high pressure? 1.5 bar or 2 bar?

what's about the 2l cross flow intake? we don't have it in france....

[fspGTD]

do you think if that intake could resist at high pressure? 1.5 bar or 2 bar?

what's about the 2l cross flow intake? we don't have it in france....

The 2l cross-flow intake has actually quite short runners if you actually look inside and measure them, they might not even be 6" from plenum to the surface where upper and lower manifolds both together... also it would be much more difficult to adapt the upper part to a 1.6lTD than an SDI manifold would be.  I'd sooner go to a 100% custom-fab mandrel-bent tubular manifold than a 2l crossflow manifold for my project...

[fspGTD]

Regarding your stock exhaust manifold, have you considered extrude honing? I don't know how much it would be, but at least you don't have to change the location of the turbo.

Peter

I doubt there would be any gains, to be honest.  I have already port-matched the exhaust manifold to the head, and stock it seems also very close to ideally matched already to the turbine inlet.  The inside walls are very smooth (yeah I did clean that up a touch, but was very nice smooth casting actually bone-stock as I recall, and also the ceramic coating probably added some additional smoothness).  The inside walls also do not narrow down at any point, so I don't see the airflow coming out of the head into the exhaust manifold seeing much of a restriction.  Enlarging the tracts inside the exhaust manifold further could actually be worse than stock, because it would mean the gasses have to narrow down more also before they can pass through the narrow slot before they hit the turbine wheel.

[TDIMeister]

PROBLEM FIXED!!!!!!!!!

Details here.

Sorry, I'd retype it, but it's almost 12:30 a.m. local time, and I've just worked on the car for 13 hours.

Take care

[caddy]

i have an intake manifold from an eurovan 2.4d 5 cylinders.

for a tdi it would work great! just need to cut one tube.
if i found some picture i will post them.

[fspGTD]

I found a study of varying intake runner lengths on a Jaguar racing 4-cylinder engine with fuel injection and non-carburated (individual throttle bodies)  Interestingly, it is also remarkably old study that it comes from a book "Scientific Design of Intake and Exhaust Systems", published by good ol' Robert Bentley in 1972.

The study shows volumetric efficiency vs rpm (equivalent to hp or torque vs rpm) curves plotted for various lengths of intake ports that were all the same (about 2") diameter.  There was no mention of cam specs on said engine, or displacement.  The lengths of the "air horns" fixed to the end of the intake manifolds were given (tested at 0", 8", 15", 21.5", and 32.5"), but the overall lengh of the induction tract between intake valve and end of intake manifold was not given.  However, using the organ pipe tuning equation and looking at the localized peaks for the various lengths, I was able to solve for the overall length of the intake tract, and I found it to vary from 4.6" to 37.1".  I also determined the reflective values observed in the plots for each resonance peak.  The lowest reflective value observed at the highest RPM for the longest induction tract length was only the 3rd RV.  4th RV was then the lowest RV for the 21.5" air horns, and 5th RV was the lowest RV for the 15" air horns.

What I was surprised to find was that even though no lower than the 3rd reflective value was used on the longest intake runners tested, the effects of the stronger reflective values was still dwarfed by a gradual, overall loss in high-rpm volumetric efficiency for the longer intake tracts (perhaps this was because the 2" intake diameter was sized too small for the engine, although that is speculation.)  This makes me think before deciding on 24" runner lengths, only looking at the organ pipe equations.  I better do some testing on my 1.6.TD before settling on a particular runner length.  I better get proficient at fabricating test intake manifolds and buy some extra material so I can make some varied designs.  Picking too long a runner length could sap high-RPM power too much, and the organ-pipe tuning equation doesn't seem to predict this important effect at all.

But there was also something positive looking at the study that is clear, that some of the very short runners were down in power across the entire RPM range.  IE: adding a little bit of length to the runners was good for quite generous (IE: up to 15% increase) in volumetric efficiency.  That kind of a gain from a custom intake manifold on my 1.6lTD, for improving my racing setup would be HUGE!

Once a runner that is ballpark-length close to ideal is found, the pipe-organ tuning equation could be useful however to predict how the local RV peaks will shift around.

[VWRacer]

Hmm, I think I have a copy of that book. I'll have to check tonight!

In the meantime, Jake, I think that the benefits of very long intake runners to a boosted engine are probably minimal. I am going to try this GTI intake on mine.

[fspGTD]

I am definitely looking forward to hearing how that works!

Edit: hey, that might even fit on your quantum fairly easily too, right?  There is adequate clearance in the area, maybe, to just bolt it on in place of the current 1.6lTD manifold?  Maybe all you'd have to do is fab some boost tubing and you'd be in business!

You might have to borrow my G-tech pro competition accelerometer before you do this, because I want you to take baseline power curve measurement and also one to compare after the manifold swap.  IMO, seeing the power vs rpm curves really helps us understand what's being effected by such a change as intake manifold swap!  IE: I'd like to see at what RPM and relative magnitude the peaks and valleys are on the power curve before vs after.  Let me know if you are interested; I could mail it to you!  I don't have any use for it at least until my Rabbit's engine is back up and running (which could be a while... potential full-motor rebuild time.)

[TDIMeister]

Jake, the reduction in VE due to extremely long runner intakes is due to the boundary layer flow loses due to the surface roughness.  That's why, as you noted,  much more has to be considered when tuning than just one parameter, such as resonant frequencies.

It should be noted that the plastic upper intake manifold has a smoother surface finish than any aluminum casting.

Here are some comments in a post related to the subject:

Unfortunately for finding a definitive, quantitave answer to whether the manifold contributes to better performance; within the last two weeks I have been using yet another chip program from what I had been using before (Ups --> ST USA). It's quite aggressive on fuelling (it was smoking like a tire-fire before last weekend's fix, but it has cleared up beautifully now, so that speaks a lot IMHO).

So, although I can't give you a quantitative number, the benchmark for comparison now would be to beat my best 1/4-mile time of 17.079 or 1/8-mile time of 10.787, both of which were achieved a week apart in late August, 2002 (my best 2003 1/8-mile time was worse at 10.87, despite .2 - .3 second better 60-ft times, reflecting a detuned chip I was running between July 2002 - July 2003 to control smoke). Hopefully I'll be able to go out and post timeslips soon...

[fspGTD]

Ahh, the boundary layer explanation ("air friction") makes sense - however, there is a picture of the Jaguar race engine with these "air trumpets" installed, and they appear to be smooth-wall metal tubing, not rough aluminum castings...  I would expect they would be pretty close in surface smoothness to your plastic SDI runners...  but, we are missing a lot of other factors in this case like engine displacement and cam timing, and how the diameter of the intake runners compares to the rate of air that is flowing through them.  IE: perhaps even at nearly 2"ID, the jaguar engine's runners are still undersized.

Anyway, I think you are definitely right that you need to consider other factors than the pipe-organ resonance equations.  It was interesting how using the equations, I was able to accurately describe the locations of each of the 3 lowest reflective value resonances in the 3 longest air horns being tested.  The observed locations of the resonances matched quite closely the calculated locations.  That tells me the equations are accurate at least for when the runners are fairly long (IE: long enough to make use of the 6th RV or lower) for comparing relatively how the resonances move by making fine-tuning changes to the lengths of the manifold runners.  Pretty cool!

The problem with measuring a modification like the SDI manifold only via 1/4 mile ET and trap speed results comparisons, is that if the manifold beefed up low-end torque but caused a loss in top-end horsepower let's say, (which I think is quite a plausible scenario I'd like to know if it's the case) the 1/4 mile ETs might show no change or a little benefit, but yet I'd like to know if the change in manifold caused this sort of change to your torque & hp vs RPM curves.  Also, since your TDI has a much lower redline I'd guessing than my IDI does, if there is a trend resulting from the the SDI manifold of hp dropping off near you motor's redline, it sure would be nice for me to be able to know that as it would probably be more acute on an IDI motor revving toward or beyond 6000rpm.