I'm going to speculate that the location of the surge line dictates boost threshold and apparent "spool" with more authority than differential rotor speeds -- the 14B car will generally feel as though it spools faster because the compressor is capable of making boost sooner under optimal conditions (assuming you aren't using an unreasonable housing that can't get enough exhaust energy to the turbine). When accelerating the turbo rotor the mass of the compressor will resist changing speeds, while creating the pressure differential (forcing the air) will deliver constant frictional loss at a constant rotor speed and increasing with speed (rather than with rate of acceleration).

What sort of concrete answer can we extract from that? No frickin' clue. I'm inclined to guess that they have an effect on rotor acceleration of the same order of magnitude.

I actually compared rotor speeds side-by-side about 6 months ago. The determination that I made about the different wheels was in reference to durability; for example the 14B is capable of around 25psi to redline on a 3/S, but the turbo probably wouldn't last long if you drove around running it to 170k RPM constantly.

While playing around with 14B turbos in non-standard exhaust housings on my car, I've been a little disappointed with the spoolup I've been experiencing. Now, I think some of my problem is due to a questionable turbo, but it got me thinking a little bit about turbo spool up.

Yeah, my 6cm^2 housings have really been doing TERRIBLY as far as spool-up goes. I think the bullseye housings would actually work BETTER for quick spool despite being bigger because they don't step down on the transition from collector to turbine scroll (if anything, they step up), and they have such a superior shape.

-Chris

 

I have also noticed a higher boost threashold than we would expect based on the flow maps. I'm not sure how the breakdown of energy goes but the energy needed to spool up the turbo shouldn't affect threashold, it should only affect lag. Also changing the size of the compressor will only marginally affect the moment of inertia of the system because the vast majority of the weight is on the turbine side. I think #2 is almost entirely responsible for the increased "spool" of the 16g. But then again are we talking about boost threshold, or lag? It would make sense if the 16g spooled (lag) as quickly or quicker given enough flow. On the other hand mass of the compressor/turbine, and RPM shouldn't play a significant role in boost threashold so the 14b should have a lower boost threashold because it requires less torque to spin the 14b compressor. Where is your boost threashold? I get very gradual boost curve, .1kg by 2400, to about .7kg (when dumps open) at just under 4k.

 

I have also noticed a higher boost threashold than we would expect based on the flow maps. I'm not sure how the breakdown of energy goes but the energy needed to spool up the turbo shouldn't affect threashold, it should only affect lag.

Obviously the energy needed to spool up the turbo WILL affect threshold if you have a very large turbine-side (as in your case and mine, and probably most TD05s). I take it you meant something else by this...

Boost threshold with these housings is pretty unbearable, I'm thinking your bullseye housings are a lot better, but STILL not quite right.


Anyhow, a real answer to the question can't come without actual curves of the shaft power requirement for driving each compressor at a given flow/PR load, and corresponding curves of shaft power delivered by the turbine at particular PR and RPM numbers.

Either way, RPM considerations will have next-to-nothing to do with boost threshold, and perhaps something to do with transient response ("lag"). Lag doesn't bother me much; high boost threshold really irritates me.

-Chris

 

Just as another data point, I finally got my JAC kit installed and the car running over the weekend. Unfortunately, boost control is escaping me at the moment, so I haven't been able to closely evaluate boost threshold, spool, and a lot of other variables too closely (too many gauges to watch). However, the few times I have nailed it hard in 2nd gear under 3k rpm, I have made over 20psi around 4k rpm's. Boost response was very gradual and linear. This is with b16g turbos in 7cm housings.

Unfortunately, I think one or both of my wastegate actuators aren't functioning properly. I welded new rods and mounting flanges in order to get compressor housing orientation more desirable, but forgot to test them off the car. As soon as I get them working, I'll do run logs with the AEM to see how early I can get full boost and what the curve looks like...

Wayne

Yeah, my 6cm^2 housings have really been doing TERRIBLY as far as spool-up goes. I think the bullseye housings would actually work BETTER for quick spool despite being bigger because they don't step down on the transition from collector to turbine scroll (if anything, they step up), and they have such a superior shape.

-Chris

 

I'd reply to each post individually, but this board is so slow it would take me my entire work day, and somehow I think my employer might frown on that.

I guess the thought behind my question went something like this:

- So far, I'm not thrilled with how slowly my 14Bs are spooling up.
- Part of why the stock 14B spools faster than the Big 16G is the smaller exhaust housing.
- I'm using an exhaust housing larger than the standard 6cm and 7cm Mitsu housings, which I expected to increase spoolup time of the 14Bs.
- In thinking about spool, it occured to me that perhaps the 14B needs the smaller 6cm housing to create enough exhaust velocity to spin the turbine quickly.
- I checked the comp flow maps, and sure enough, the 14B has to spin VERY fast in relation to other turbos.
- Since the big 16G spins ~20% slower than the 14B to create comprable boost, perhaps it would actually spool faster (fromt a boost standpoint, not an RPM standpoint) than the 14B in these large exhaust housings.

Thus, my question about intertia vs aerodynamic drag. If the 14B and 16G are very similar in terms of rotating mass, and overcoming the intertia of the wheels is the primary task of the exhaust energy, then I would expect both turbos to have a similar RPM curve. That is, given X exhaust passing through the turbine, I would expect both turbos to achieve Y RPMs. If this is the case, then the 16G would make more boost with at that RPM level, and thus boost would come on faster than the 14B.

On the other hand, if aerodynamic drag of the compressor wheel sucks a significant amount of the energy harnessed by the turbine, then the above theory goes out the window...or the wastegate, I suppose. My gut tells me that drag probably does play a significant part. But, there are always those cases where the math proves your gut wrong, and had an inkling this might be one of them.

Since doing all this musing, I discovered that the wastegate on my rear turbo was being partially held open. This certainly was not helping my spoolup! I've fixed the problem, and I'll do some more experimenting with spoolup later this week.


On spool vs. boost threshold: I'm not sure we're all on the same page about what these terms mean. I think we're all using "spool" the same way (ie how quickly the wheels spin up, and thus how quickly boost is available). But I'm not sure about "boost threshold". I know what it means to me: given sufficient exahust energy to spin the turbine (ie no spoolup time), the maxiumum boost the turbo can create. What are you guys meaning when you use the term?

 

To clear this up:

On spool vs. boost threshold: I'm not sure we're all on the same page about what these terms mean. I think we're all using "spool" the same way (ie how quickly the wheels spin up, and thus how quickly boost is available). But I'm not sure about "boost threshold". I know what it means to me: given sufficient exahust energy to spin the turbine (ie no spoolup time), the maxiumum boost the turbo can create. What are you guys meaning when you use the term?

Boost threshold is the point (engine speed) at which you are first able to reach the desired set boost. If your set boost is 1bar and you can reach that at 3000rpm in third gear IGNORING time it takes to reach that pressure, only attending engine speed at which that pressure is reached (a quasi-steady-state situation), your boost threshold is 3k. You should be able to reach that boost by that RPM in ANY gear by neglecting transient effects.

In other words if you can hit 15psi at 3k in third, you can have the same amount in first or second by applying the brakes so as to maintain engine speed at 3k. This "boost threshold" represents the ability of the compressor to motivate air, considering the power it receives from the turbine in an operating environment.

Transient response is another way of referring to the time delay between opening the throttle and coming on-boost -- this varies tremendously with engine speed, both above and below the boost threshold. A comparison between two engines could be made by observing the rate of boost rise (with a logger) when the throttle is suddenly opened at a particular RPM (keep the gear the same and as high as possible for consistency).

"spool" is a nebulous term that people variously use to describe threshold AND transient response.

-Chris

 

To clear this up:

Boost threshold is the point (engine speed) at which you are first able to reach the desired set boost. If your set boost is 1bar and you can reach that at 3000rpm in third gear IGNORING time it takes to reach that pressure, only attending engine speed at which that pressure is reached (a quasi-steady-state situation), your boost threshold is 3k. You should be able to reach that boost by that RPM in ANY gear by neglecting transient effects.

In other words if you can hit 15psi at 3k in third, you can have the same amount in first or second by applying the brakes so as to maintain engine speed at 3k. This "boost threshold" represents the ability of the compressor to motivate air, considering the power it receives from the turbine in an operating environment.

Transient response is another way of referring to the time delay between opening the throttle and coming on-boost -- this varies tremendously with engine speed, both above and below the boost threshold. A comparison between two engines could be made by observing the rate of boost rise (with a logger) when the throttle is suddenly opened at a particular RPM (keep the gear the same and as high as possible for consistency).

"spool" is a nebulous term that people variously use to describe threshold AND transient response.

-Chris

Gotcha. Boost threshold is kind of a "best case" scenario for the lowest engine speed at which a given boost level can be achieved. I was thinking of it as more of a "given infinite exhaust" term, but your definition is more useful.

In my lexicon, spool == transient response. But, I know what you mean about people using it to mean other things.

- Brian

 

- So far, I'm not thrilled with how slowly my 14Bs are spooling up.
- Part of why the stock 14B spools faster than the Big 16G is the smaller exhaust housing.
- I'm using an exhaust housing larger than the standard 6cm and 7cm Mitsu housings, which I expected to increase spoolup time of the 14Bs.
- In thinking about spool, it occured to me that perhaps the 14B needs the smaller 6cm housing to create enough exhaust velocity to spin the turbine quickly.
- I checked the comp flow maps, and sure enough, the 14B has to spin VERY fast in relation to other turbos.
- Since the big 16G spins ~20% slower than the 14B to create comprable boost, perhaps it would actually spool faster (fromt a boost standpoint, not an RPM standpoint) than the 14B in these large exhaust housings.

I'm not entirely convinced the 6cm^2 housing is any better with the JAC headers, in fact I think it may be worse. The outlet on the JAC headers is a very reasonable 2-1/4" diameter, while the inlet on a 6cm^2 housing is 4/10ths of an inch smaller; that HUGE step down is likely responsible for disruption in exhaust flow and reversion at the turbine inlet flange. A properly ported 7cm^2 housing starts at the same size as the JAC flange and tapers into the scroll, while an unmodified Bullseye housing starts out at the same size (or even a hair larger) than the JAC flange and moves into the scroll with a gradual conical taper (much nicer than the 7cm^2 one if you've compared the two).

I would think the Bullseye housing would be superior in every way to the 6cm^2 housing, and still spool faster than a 7cm^2 housing when paired with the JAC kit. There IS a loss in velocity using such large cross-sections in the exhaust path, but you certainly aren't going to make up for it by putting an abrubt and large step-down into a restrictive scroll.

If JAC Engineering's kit used slightly smaller tubes and a smaller collector outlet, the 6cm^2 housing would be a better choice. I think the current collector outlet size on the JAC headers is just right for max performance applications (even if the collectors themselves leave something to be desired).

-Chris

 

Thus, my question about intertia vs aerodynamic drag. If the 14B and 16G are very similar in terms of rotating mass, and overcoming the intertia of the wheels is the primary task of the exhaust energy, then I would expect both turbos to have a similar RPM curve. That is, given X exhaust passing through the turbine, I would expect both turbos to achieve Y RPMs. If this is the case, then the 16G would make more boost with at that RPM level, and thus boost would come on faster than the 14B.

The answer here may involve a relatively easy-to-understand fluid mechanical concept:

The maximum power delivered to a turbine occurs when the turbine blades are moving at 1/2 the velocity of the impinging jet. If we could get an idea of average exhaust velocity at the outlet flange (I have no clue how we would do this), we could determine what ballpark we're operating in at 95-170k RPM rotor speed, and approximately what speed peak power is delivered to the turbine (and where this falls on the compressor's maps).

-Chris

 

I'm not entirely convinced the 6cm^2 housing is any better with the JAC headers, in fact I think it may be worse. The outlet on the JAC headers is a very reasonable 2-1/4" diameter, while the inlet on a 6cm^2 housing is 4/10ths of an inch smaller; that HUGE step down is likely responsible for disruption in exhaust flow and reversion at the turbine inlet flange. A properly ported 7cm^2 housing starts at the same size as the JAC flange and tapers into the scroll, while an unmodified Bullseye housing starts out at the same size (or even a hair larger) than the JAC flange and moves into the scroll with a gradual conical taper (much nicer than the 7cm^2 one if you've compared the two).

I would think the Bullseye housing would be superior in every way to the 6cm^2 housing, and still spool faster than a 7cm^2 housing when paired with the JAC kit. There IS a loss in velocity using such large cross-sections in the exhaust path, but you certainly aren't going to make up for it by putting an abrubt and large step-down into a restrictive scroll.

If JAC Engineering's kit used slightly smaller tubes and a smaller collector outlet, the 6cm^2 housing would be a better choice. I think the current collector outlet size on the JAC headers is just right for max performance applications (even if the collectors themselves leave something to be desired).

-Chris

My speculations were based purely on the turbine scroll area - I wasn't thinking about the effects exhaust housing entrance at all.

That being said, you're absolutely right. Putting a 6cm housing on the JAC headers without porting the entrance to match the header outlet, and porting out the step, is like committing spoolup suicide. I've had all three housings in my hands (actually, I have a couple of 7cm housings if anyone needs some!), and the bullseye housing is absolutely the most well-designed, and the best match for the headers.

- Brian

 

Ok, I did a basic analysis of gas flow in the engine to give a ballpark number for exhaust gas velocity. I made a number of assumptions and did the analysis for 7k on Jeff's model of an engine.

At 7k, one bank at ~20psig consumes ~400cfm (ambient temp and press), or approximately 30lb/min of air. I assumed the exhaust gas at the collector to be at 850 degrees Celsius, and about 40psig (twice intake pressure). At this temperature and pressure, it is approximately the same density as the ambient air as per the ideal gas law; I have neglected the addition of fuel and the composition of the end gas, I'm not sure how much this changes my analysis but I was looking for a ballpark figure.

So about 400-433cfm is flowing through the exhaust of one bank, or approximately 7cfs. 7cfs of air through a 2.2" diameter circle (.0276 sq ft area) would move at 253ft/s. Assuming there is no change in density (so a constant volume flow) from the collector outlet to the turbine inlet, 7cfs of air through a 7cm^2 area (0.007535 sq ft) would move at 929ft/s.

The inducer of a TD05H turbine wheel is 2.2" -- I am assuming that the exhaust at the 7cm^2 area is impinging upon the wheel tangent to its rotation, and that it impinges upon this diameter, not a smaller one. The circumference of a 2.2" wheel is about 0.5759 feet, and 929ft/sec divided by 0.5758ft would be 1613rps or 96,787rpm -- so about 100,000rpm.

Now I have made a LOT of assumptions, but I don't expect they cause more than ten percent variation on the answer, so we have a reasonable ballpark for average velocity flow at the turbine inlet. What are the problems with this perspective?

1. While 929ft/s might be the average exhaust velocity at redline and 20lb boost, it is certainly not the PEAK velocity. Theoretically if the exhaust never moved faster than 929ft/s, the turbo would NEVER make boost because the exhaust could not turn the turbine wheel fast enough. Obviously this is false -- it must be moving faster than that for some portion of the time (though the mean flow rate I calculated is much lower than that peak number).

2. Assumptions were made about the exhaust gas doing work to the turbine wheel. The design of the wheel has been entirely neglected, including the axial component of this axial-radial turbine. Admittedly this is irrelevant for a simple velocity analysis as I have made.

3. I have not actually measured the pressure and temperature in my headers at 20lbs of boost. While I plan to use an EGT and pressure gage on the headers at some point, I took what I thought to be reasonable estimates for exhaust backpressure and a thoroughly tested common exhaust gas temperature for this estimate. Not only that, but because gases are compressible, and pressures as the exhaust valve opens are VERY high, I have no clue as to what the changing density and velocity in the exhaust manifold would be.

-Chris

 

Unfortunately, I don't know what to do with that kind of an estimate either. My understanding of fluid dynamics is rudamentary at best. I bet someone's written some good simulation software for things like this...but I also bet that software is obscenely expensive.

- Brian