Mounted the maf sensor right behind the air filter. Hot wire type sensor.

I thought that there is the ideal place- for the sensor.

AlpineRAM

 

Alpine--

Good post!

I think you are being VERY conservative with your air estimates. You can use an AFR of 22:1 and a BSFC of .4 and get good results. In the spirit of efficiency, as you posted, you only need so much air-- so shooting for 120lb when you only need 90-100 or so will really hurt efficiency.

Maximum mass flow at 31psi is about what we've speculated for a long time but now we have data. Thanks! Was this 31psi at the intake manifold or at the compressor outlet? Do you know what kind of pressure loss you were seeing from the compressor outlet to the intake manifold?

If you use a tiny turbo like an HY35 for the secondary, then it's difficult to get very effective compounding because you end up having to bypass so much exhaust. Say you used an HY35 for quick spoolup. You'd have to pop the WG on it open at about 20psi because it's already choking from the tiny turbine side.

This means we are getting about 1.5 PR from the top charger. Now, we have to run our big turbo at a lot higher PR to get the total pressure ratio the engine requires. Not only this, but we lose efficiency because the not all of the hot exhaust is passing through two turbines-- but rather a lot of it is bypassing the tiny top turbo and going through one turbine only.

Will we have enough drive energy to push that large primary to a 3:1 or higher PR once the wastegate onthe secondary opens? This is the problem with twins that use a huge size difference between turbos. The small turbo has to be big enough to get the large turbo fully into it's operating range.

The wastegateing is the real trick, imo. The pressure signals to the wastegate(s) need to only "see" the pressure ratio for that stage.

So here's how I figure twins calculation, using a 550hp goal for an HO truck:

-- I want a turbo that can make 2.5:1 PR with reasonable efficiency on top, and I want it to spool pretty quickly. A stock HX35 with a larger turbine housing would work. In the flow range we are looking for, you have to go with an external gate. The sizing of the gate should be chosen based on the "bypass ratio" of the gas flow (how much goes out the gate when it's fully open vs through the small turbo). Since we're going with an external gate, we'll use a 16cm housing on top.

-- The HX35 with a 16 housing can efficiently make 2.5:1 PR, or about 22psi of boost at sea level. This will keep is in the middle of the map and discharge temps will be optimized. We hit 22psi, the wastegate will start to crack. At this point the primary must already be making positive pressure in the crosspipe. A GT4202 will already be making positive pressure with the .85 AR housing.

How do we know this? Well, at a PR of 2.5 the engine is "appearing" to be much larger than it is, based on air flow. With no boost, our 5.9 is flowing 95 CFM per thousand rpm. So at 1500 rpm, we're looking at about 135cfm.

But this is when it's drawing in air at 14.7 psia. What if we doubled the inlet pressure? Well, now we're flowing double the CFM! If we triple it, we've tripled the CFM. In other words, the engine "appears" to be double or three times the size that it really is.

Now this assumes we've increased pressure without increasing temperature, which we can't do. Since our compressor is about 70% efficient, we will only see 70% of the increase in pressure show up as an increase in density. So if we run the PR at 2.5:1 the airflow that the small turbo is swallowing is only as if we were running a 1.75:1 PR with no heating.

Let's convert this to mass, as we probably should have been doing all along. Figure .0765 lb per cubic foot of air at sea level. So our engine at 2000 RPM and 2.5PR is swallowing air at a rate of 332.5 CFM, or 25.5 LB/min of air.

So now we ask the question- will I have my primary spooled up? In our case, the turbo is the GT4202, and the turbine map tells us that that at 35lb/min corrected gas flow (with the 1.01 A/R housing), the turbo is spooled up. But the key word here is CORRECTED gas flow.

Our engine is swallowing 25.5lb of air that's relatively cool (say 100º). Then it's burning fuel and putting out that same 25.5 lb/min of air (ignoring the mass of the fuel burned) and spitting it out at MUCH higher temps. This high temperature is the key here, because it's ENERGY that drives the turbine-- not mass. Energy comes from mass, temperature, velocity, and other sources we can reasonably ignore.

The bottom line is that 25.5lb or gasflow at 1200 degrees or so is MUCH more in terms of energy than 35lb of mass flow at room temperature.

We could do those calculations too, and we could actually figure out where the primary will start spooling and so forth. In this particular case, though, we could say that our GT4202 is well spooled and making good boost. But that' completely depends on what "corrected" means. Corrected to what temperature?

Anyway, we don't know how much boost it's making, or even capable of making.

This is where to mass flow ratios of the two turbos are useful to compare. If the small turbo can swallow 25.5 lb/min at a PR of 1 (atmospheric pressure), then how much air does it take to supply it at a PR of 2:1? Or 3:1?

As we did before, we know have to account for inefficiency. The inefficiency doesn't show up as a higher mass flow requirement, so be careful of trying to say that "x mass flow causes Y pressure ratio" You have to do PR first and re-iterate. Remember, flow is caused by a pressure difference, and the rate of flow is a function of 1) the differential pressure, and 2) restrictions. So we do PR first, because PR is the cause, and mass flow follows.

Because of heating of compressed air, we're going to have to run that primary over a pressure ratio defined by (ideal PR/efficiency). In other words, at 70% efficiency, we have 3/.7, or 4.28.

So in this particular scenario, we need a compressor that can flow around 75lb/min at a PR of over 4.2:1 because we have no between-stage intercooling. This is the ragged edge of the compressor map, and far from ideal. It would be better if we could drop the PR down to 3:1 or so, but that's not happening unless we can get that temp down. (intercooling).


**break**

I'm planning to build twins that will support 550hp with no smoke and low EGT. By my calculations, running a GT4202 at a PR of 3:1 and the top charger at a PR of 2.5:1 will do this more than adequately even with no intercooling-- just the factory CAC.

If I use the stock charger as the top turbo, here's how it plays out at 2000rpm. In this case, the big charger is taking in 100º air, compressing it 3:1 and discharging it at 368º. The stock charger takes in this 368º air, compresses it at a ratio of 2.5:1, and discharges the air at 681º!

Now we can only support the 450hp at 2K rpm if we can cool this 681º down to at least 450º. The cooler we can get the air, the better-- much better.

The point is that this set of twins will spool very well and cool 450hp at 2K rpm, and up to 600 hp at higher rpm.

Even in 100º weather. In Denver elevation.

Now I just have to build the darn thing and put my money where my mouth is.

I'd love to upgrade the top charger to a Garrett also (GT3782, stg 2), but my budget probably won't allow it. That stupid Spearco intercooler is gonna run almost $1500 alone!

Justin

 

Justin- My thoughts about getting a very small secondary come from the following considerations- In our pickup trucks the engine will only tick over most of the time and will see short bursts of WOT - I can't imagine a situation where I need a constant 500hp @ 2k rpm- the vehicle will accelerate with this kind of power, so this is only a transitional phase. For those short bursts we need much more air than while cruising, but you wanna be prepared. (vis pacem, para bellum) - So my consideration is to use the small charger "sucking" through the big one to stay in it's happy region, and still prespool the big one, it won't need to be fully spooled. My assumption is to open the WG very early on the secondary to achieve the main pumping work at WOT with the big primary, and driving the secondary just enough that it isn't a choke point for the fresh air to come in- all of this with an extremely free flowing gate on the secondary- I don't really think we need to compound compress too much for the amount of air needed- the GT4202 will supply enough air for keeping short bursts of 500 hp cool enough for my taste- I'm considering building the secondary more of a straight pipe when the WG is open and a diversion through the turbine when it's closed. (as a philosophy- because when cruising a little more resistance doesn't matter too much, but under WOT I'd like to have about 66% of the exhaust gas in the primaries turbine, and only 33% through the secondaries turbine)
I think that the "arms race" for more hp is a little misguided- I'd like to see lots of HP that are usable right now, instead of using my calendar to make an appointment for the time I've got up to boost - this is my personal preference since I don't drag race and don't do sledpulling. To me it's important to be able to accelerate quickly between the turns on mountain roads, to pass the other vehicles without causing solar eclipses etc- so considering spool times of the classic twins or big singles isn't what I'm looking for. I have driven an HY equipped truck and the really quick spool made more usable hp than the modded HX I have in my situation. - This lead to the idea of building something that should work well on the mountains but still have enough oomph on the highway to be a nuisance to Porsche drivers.

I used the MAF sensor right behind the air filter, the pressure sensor was in the intake horn and I didn't measure temperatures. I also don't have data on the pressure loss across the CAC.
I think that intercooling between the stages of compound compressing is to be avoided as long as possible, as long as the charge air temps don't damage the secondary I think we are better off with hot air that flows easier, less volume to introduce lag, and a bigger delta T for the charge air cooler. ( Assuming that we must get rid of the energy from the intercooler)- I think it could be beneficial to utilize an air to water aftercooler after the stock CAC, plumbed into the radiator return line to get the coldest water possible.
Your calculation of corrected flow has some flaws IMO- the engine has a lot of scavenging as long as the boost is higher than the exhaust backpressure- so air consumption of the engine is also going down with rising boost since the backpressure is also rising. In some place on the operating map you have a reversal of situations- from an engine that does scavenging to one that actually reuses exhaust gas. (Naturally this is the wrong way round since egr happens under WOT..)
I think that we'll need not very much boost for high hp as long as we keep the proper pressure ratio between intake and exhaust.

In my situation the twins project will stay theory for some time to come, $$$ or to be precise €€€ being the reason.

I hope we can keep this discussion running. It would be very interesting to see mass air flow with different cams under certain situations- especially MAF while cruising for stock, stock with a different turbo or twins, aftermarket cams for those setups etc.


AlpineRAM

 

Jeez, and I thought I could just use the new ricer trick for more boost

 

It's important to sample the intake air with sound methodology - for example, pre-CAC temp and airhorn psi alone aren't particularly useful for data acquisition.

Just to avoid confusion, it's better when you guys clarify absolute or gauge pressure when talking about ambients, TIP, or compressor discharge - sometimes it seems as though they are interchanged within the same post.

I know it's easy to forget that the metric side of the pond uses different assumptions - I've been guilty of it!

Keep up the thoughtful posts!

 

There should be a Darwin award for DIY engineers!

 

Alpine, I'm with you. I think we are like-minded in what we want twins to do.

Perhaps we should consider the twins thing in three phases:

1) Low rpm. (say 1500). Here's we want to have enough drive pressure to where the small turbo is already making some kind of positive pressure. As long as it is making SOME boost, we will have dramatically improved response over none at all.

2) Medium rpm (say 2000). Here we want to have instantaneous response available, We also need compounding because the PR is going to be be very high. We wa

3) High RPM (2500+). Now, we don't need as much compounding because the RPMs have risen enough to move the air we need without cramming. Hence, we don't need as much "compounding" effect. For example, we can make 450hp at 2000rpm with a PR of 4.5, but it only takes a PR of 4.4 to make 550hp at 2700 rpm. The RPM helps alleviate the need for more PR.


So what we can do is think of this now in terms of turbo workload. Phase 1 is basically all small turbo. Phase 2 is both, and phase 3 is mostly large turbo.

We'll have to manipulate turbo workload via wastegating. Because on the top end we'll be bypassing a LOT of flow, we'll need a huge gate-- like a 60mm Tial or a t-netics NewGen HP or such. We need a wastegate that has two pressure references so it will "see" a pressure differential, not just compare a pressure signal to atmospheric pressure.

We do this because we don't just want the WG to pop at 25psi or so. What we REALLY want is for it to pop when the top turbo is making 25psi MORE THAN it's receiving from the big turbo. So we need two boost references- the crosstube and secondary compressor discharge.

This allows us to shift proportionally more and more workload to the big turbo as total boost pressures rise. This relieves a lot of the restriction a small turbo would pose because we can bypass more gas.

For example, just before the wg pops, the top charger is operating at a PR of ~2.7-- it's taking in 14.7psia and spitting out air at 39.7psia.

But at high boost levels, the proportional PR the small turbo is making drops. Let's say that the big charger is feeding it with a PR of 4:1-- or about 44psig boost. Now when our top charger adds its 25psig, we have 69psig boost-- but the PR of the top charger has fallen because it's "multiplying" 58.7psia to make 83.7psia-- which is a PR of only ~1.4.

The trick is finding a way to configure the WG to only see the pressure differential-- and I'm not sure how to do this. I don't know how most externals work. I'm under the impression that there's a nut on top that adjusts spring preload, and then there's also some kind of pressure signal. The wastegate we'd need to do this would have two pressure ports-- one on either side of the WG diaphragm. So we just drop in a 20psi spring or so, connect the "spring opposed" side to the secondary boost signal, and connect the "spring aided" side to the crosstube.

Finally, we'd need a wastegate on the big charger. This would be much simpler, since we could just set it up to pop at whatever PR we want from the primary (say 4:1), and dump the gases into a blowdown tube-- no need to plumb it back into the exhaust, imo. After all, the wg on the primary would only be open for very brief moments.

Anyone know of a wastegate that has the "dual boost" references that I mentioned? Do we think a dumb wastegate is fine if you use a smart boost controller to it's lone boost port?

IMO- the wastegate tuning is a pivotal element of getting twins to function right-- second only to turbo and housing selection.


JMO

 

Interesting. While I can offer nothing in terms of the theory here, I can tell you my real time observations of my particular set. My sig says what I have. I have boost measured at the intake horn and I also have boost measured at the cold pipe going from primary to secondary. First, the hx35 has an enlarged wastegate. Not sure on actual size but you can stick your whole thumb in it and have a little room left over. The wastegate is set to crack at 25psi and is fully open at 28psi. Now the following are just rough guesstimates and vary by conditions but at cruising speed--65-- I see a total of 3-5psi boost. All from top turbo. When rolling into the throttle, I'll see aproximately 1/4 to 1/3 total boost from lower turbo. For instance, with 15 psi at the intake horn, the BHT3B, lower turbo, is putting out 5-6psi. at 25psi total boost, the lower is supplying about 10 psi. By the time I get to 40 psi total, the lower is supplying 20-23psi of boost.
Now about some of the air temps discussed elsewhere, I can tell you that after a WOT run, you cannot touch the turbo side of the intercooler as it is very HOT! But you can touch the opposite side. It is very warm, but you can touch it. So the intercooler is doing its job.
Hope that helps the discussion.

 

FWIW-- the lower turbo is not wastegated. It never sees anything close to an operating limit. However, there are blowoff valves on the tube going from top turbo to intercooler set to about 50psi to save the headgasket.

 

I'm not a fan of those hot water heater blowoffs, because they are wasteful. If you only need 35psi from the bottom turbo, why make more? That extra boost isn't free, it came at the price of drive pressure.

Far better, imo to wastegate that big charger and lower the drive pressure.

 

FWIW... not disagreeing with you, you have a point. But I don't get to 50psi anyway, not enough fuel....yet

 

I am new to the "twins/compound" thing but would it make sense to operate the internal wastgate from the boost in the cold pipe? I am guessing that would keep boost up until the primary is lit. Comments???

 

Yes and no. With the wastegate setup that way, you can't control the boost relationship between the two turbos past the opening point. If you run your top charger off the cold pipe, then once the gate pops wide open, the only flow regulation is the restriction of the wastegate vs the restriction of the small turbo's hot side. In other words, once the bottom charger reaches the WG set point, you have no control over the "bypass ratio" except sizing.

You can do it this way, but it's imprecise and harder to tune. It's a lot tougher to swap out turbine housings and wastegates than it is to just tweak a wastegate setting.

To my thinking, you want to control the top turbo's wastegate based on the workload the TOP turbo is doing, and you want to control the bottom turbo's gate based on the workload the bottom turbo is doing.


You are correct, though, that controlling the top wg with the cold pipe pressure would prevent the WG cracking until the primary lit-- but there's life after the primary. We want to control both turbos across their entire operating range, instead of just until primary spoolup.

I've even heard of wastegates being run off drive pressure signals-- which is intriguing, but I'm not sure of the value of it.

If you can't run the top turbo with a "dual port" wastegate, then the cold pipe is probably your second best option-- assuming you've done well selecting turbo and WG sizes.

JH

 

boy this went way past my ??

 

Yeah, but it's been a pretty good hijack!