[Atg]

Someone mentioned original application for he451:

PEGASUS L SERIES ISC8.3 - QSC8.3 - 6C - ISL9 Holset HE451

I have some part numbers now for the wheels and the turbine housing, but still need to measure it.

 

[Atg]

This is probably the turbo from the ISL G developed in conjunction with Westport for CNG use.

http://www.cumminswestport.com/models/isl-g

 

[Atg]

These guys think a 16cm housing is approx 1.13 a/r

http://turbolabofamerica.com/turbine-housings/

 

[Atg]

Ok I wedged some modeling clay down there and fired up a photo editor. One half of the inlet is 762 mm2. So approx 1524 mm2 total at the inner end of the ramp.

Turbine wheel is 64mm small and 74.87mm large diameter.

Will have to estimate a scroll radius later.

 

[Atg]

Ok I wedged some modeling clay down there and fired up a photo editor. One half of the inlet is 762 mm2. So approx 1524 mm2 total at the inner end of the ramp.

Turbine wheel is 64mm small and 74.87mm large diameter.

Will have to estimate a scroll radius later.

Short version:

Holset He451
Part numbers 2837788, 2837789
Original turbo on cummins ISL9 etc

Compressor Inducer 64mm
Compressor Exducer 93mm
Compressor blades: 7+7

Turbine Exducer 64mm
Turbine Inducer: 75mm
Turbine blades: 12

Turbine housing radius: 6.95cm
Turbine cross sectional area: 16cm2
Turbine A/R: 0.9 inches
Twin scroll


Long version:

From the above I assume the true area at the end of the tongue is actually 16cm, and that my modeling clay method is off a bit.

Compressor inducer was a bit more difficult to measure, as it has the odd number of blades at the inducer. Basically, I got a triangle from the tips of blades 1,3 and 5 of a seven blade config, with tip to tip measurements of 50,50 and 62.6. If anyone has an easy way of fitting a circle to those points please let me know, or I will do some math later today to figure out the diameter of the circumcircle (haven't found that online calculator yet). It is probably 64mm because if I put the sides at 5, 5, and 6.2 on this java thingy http://www.mathopenref.com/trianglecircumcircle.html it comes out at 6.38cm diameter, but the 6.2 side should really be 6.26, which makes the diameter a bit bigger than 6.38.

Compressor exducer I measured at 92.9. This seems dramatically larger than the inducer, which apparently moves lots of air, but less efficiently than lower trim wheels. It has 14 blades down low. It has machining marks on it, so I assume it is billet, unless the marks are transferred from the casting mold.

Area ratio A/R on the turbine seems to be 16cm2 divided by 7cm, or 2.28. I measured it at 6.95cm so it could be a bit larger than that actually. If we use my actual measurement of 1524mm2 and 69.5mm radius, we get 2.19 after converting to cm. I tend to trust the 16cm2 value stamped in the housing, so 2.3 is probably closer to the truth. Apparently this is equivalent to 0.9 inches, which is the conventional unit.

Turbine has twelve blades on it, if I didn't mention that already.

I take all this to mean that the turbine wheel is larger than the guy on the supra forum indicated, which should be good for backpressure. The turbine housing A/R seems medium-ish.

The compressor also appears high trim. But it should flow a bit more on the top end, where we were bumping into a flow limit at 6000rpm I believe.

I would put some photos in, but I cannot figure out how that works just now.

 

[Dougal]

The A/R you've measured is in cm. Most use inches which is 0.9.

 

[Atg]

The A/R you've measured is in cm. Most use inches which is 0.9.

OK thanks Dougal. I've corrected it in the previous post.

 

[Atg]

Aha!

Found this gem: The diameter of the circumcircle can be computed as the length of any side of the triangle, divided by the sine of the opposite angle.

from here: http://math.wikia.com/wiki/Circumscribed_circle

Anyway, it requires solving for the angle opposite a side. So we have Vertices A, B and C, angles alpha, beta and gamma, sides a, b, c.

By law of cosines, alpha = acos (b^2 + c^2 - a^2) divided by 2bc.
= acos [2500+3918.76-2500 divided by 2*50*62.6]
=acos 3918.76/6260
= acos 0.63
= 51.24 degrees

so a=50, alpha =51.23
circumdiameter = 50/sin 51.23
= 50/ 0.7727
=64.127mm is the diameter of the compressor inducer from measuring distances between three of the seven blade tips of 50, 50 and 62.6mm

 

[Dougal]

Turbine specs match an HX40-16cm pretty close, it's about 20% bigger (flow wise) than an HX35-12cm turbine.

So it's smaller than I initially thought (26cm became A/R 1.5 inches). That website with a list of cm-A/R conversions is only true for some MHI TD05 turbos. It's wrong for everything else as the radius differs from their assumptions.

I think what we're seeing with all these newer generation Holset turbos is bigger compressors and tighter turbines (compared to compressor size) to give a turbo that naturally leans out and cleans up a diesel engine. So you've basically got an HX55 compressor with a HX40 turbine attached.

 

[Atg]

I wonder if an 25cm hx40 hot side would bolt on. That would help backpressure and shift things toward the original 451 estimate by Dougal.

 

[Atg]

Apparently to get big exhaust housings like 28cm etc, the hx55 is best.

Hx52 is typically an hx50 turbine wheel with an hx55 compressor, so might be better turbine flow than 451, but housing is still 16cm usually.

Hx55 have the range of huge turbine housings.

Above info from here http://www.compressorracing.com/our-turbos.html

If I can find a core hx40 with a 25cm housing, I have a feeling it may bolt up to the 451. So that would probably be the cheapest way to improve exhaust flow.

Conceptually I have trouble understanding how turbine exducer size and housing sizes each contribute to backpressure.

 

[Atg]

I can always put an external gate on the LP turbo also, to keep the pressure ratios split more evenly. I don't really want to mess around with trying to bypass the HP compressor. I guess it just depends what the map looks like on an he451/hx55 compressor, e.g. what pressure ratio it likes at the target flow rate. And also what it looks like from a backpressure standpoint.

The 120cc Calc above for hx50 and he451 assumed a smaller turbine wheel, but a much larger turbine housing, than I actually have. So now there is a big 16cm fixed obstruction limiting what can go thru the LP turbine, but a larger wheel. I assume this will be a net increase in backpressure.

Either I waste gate the 451, or I look for an hx50 with a larger turbine wheel and housing to lower backpressure.

 

[Atg]

HX30 Super and HE451. I used 26cm for the HE451 housing which comes out I think around 1.5 A/R.

Big disclaimer for these results. I can't guarantee any of these components will last when used in this manner. The HE451 will likely be fine, but it's going to be a hard life for the HX30 Super and also for the engine, the gearbox, the driver and everything else.

I used 65psi boost and that came out to more like 170cc for the midrange with 155cc at 5500rpm. 17:1 A/F was used throughout.

6000rpm went off the map a little bit for the big turbo, so even though you can spin it that quick, you're well out of efficiency and power making territory.

Because the HE451 isn't wastegated it tends to gain boost with rpm, peaking at PR 2.73 (25psi) at about 5500rpm.
Max boost is still looking like it arrives a bit past 2000rpm. This is going to be scary to drive!
Torque is over 1000Nm from just past 2000rpm to 4500rpm.
Power will peak around 4000rpm where 400kw is expected (540hp).

Drive pressure will top 120psi by 6000rpm, but at 4000 is going to be around 80psi and at 2000rpm drive is below boost.

The HX30 super is on the map the whole time (based off my guesstimate of an HX30 Super Map) with ~35 lb/min max and the HE451 runs just off the map at 6000rpm. This is ~80 lb/min.
But by 6000rpm you're down to ~265kw anyway so you should have changed up a gear ages ago.

Maybe this will be fine with the two turbos I have if I run it at high boost. My LP compressor is bigger than assumed above, which is good. So it all depends upon the backpressure. I think it should never exceed 2.5x the boost, so it really depends upon whether this 16cm LP housing will flow well enough to keep emp below 160psi at 5500 rpm or so.

 

[Atg]

hx52 16cm at 160cc om606

https://www.dropbox.com/s/cel7evzvpkwps1p/2016-03-03 08.57.30.jpg?dl=0

 

[Dougal]

https://www.dropbox.com/s/cel7evzvpkwps1p/2016-03-03 08.57.30.jpg?dl=0

So what was used on that engine? Looks peaky!


I haven't had any time to run more calcs.

 

[Atg]

Sounds like it is time to ply you with turbo parts.

Hx52 16cm 160cc; not positive on fueling level but I think that is correct. Sorry I meant to post that info above with the graph.

I think his turbine is limiting the top end. This is because it looks just like a graph from the book you mentioned which shows what undersized turbine does to the power curve. Hx52 compressor should be large enough. So it must be the 16cm housing. It is enough power for sure, but perhaps not ideal turbine housing size for my hx40 sized turbine and a compound setup.

Seems like an hx55 22cm would be lots better for a compound with hx30 and my 150cc pump. They are cheaper than the 451 I bought. But then hx30 6cm seems small, unless I put more PR on the LP turbo and intercooler that stage rather than the HP stage.


The hx30 super showed up from China last week. Thanks Alcaid!

So what was used on that engine? Looks peaky!


I haven't had any time to run more calcs.

 

[Dougal]

So I've got a few spare minutes to run some calcs. So is the intent to run the 16cm 451 ungated or with an external gate?

I'll do a run with the 451-16cm ungated at 120cc and see how it looks.

*edit*
So I've run some ungated calcs and this is looking surprisingly good. You'll need a huge wastegate to bypass the HX30 pretty much completely from 4000rpm to 6000rpm.
Power should be around 350kw (crank).
Drive pressure is below boost from 4000-6000rpm.
The 451 runs straight up the middle of the compressor map.

You may or may not find the HX30 compressor to be a restriction at higher rpm. It's too close to call A flap valve to bypass it (with a wastegate actuator set to ~35psi triggered off the HE451) would cure that.

 

[Atg]

Thanks Dougal. I have been going back and forth on the big turbo selection question.

Eventually I decided to just bolt it up and see how it goes. A guy stamsas in Norway has hx35/hx52 compound on a 606 and there is a dyno run on facebook which has both of them spooling quite late. Another guy Hans Persson in Sweden run straight hx52 but very late spool. All hx52 have 16cm housing, but larger turbine wheel than he451. So I think the smaller wheel on mine with similar housing cm should be about right.

I had not considered the flap compressor bypass until seeing blacksmoke racing on facebook with he221w under hx50 sequential. It has a gizmo like that. Somehow. Does the hx30 super contribute anything at 4000-6000? It is still presumably cranking away at 40psi or whatever. I did get the super so that should help.

Anyway now just have to either build the manifold or get someone working on it.

Cheers,
Karl

So I've got a few spare minutes to run some calcs. So is the intent to run the 16cm 451 ungated or with an external gate?

I'll do a run with the 451-16cm ungated at 120cc and see how it looks.

*edit*
So I've run some ungated calcs and this is looking surprisingly good. You'll need a huge wastegate to bypass the HX30 pretty much completely from 4000rpm to 6000rpm.
Power should be around 350kw (crank).
Drive pressure is below boost from 4000-6000rpm.
The 451 runs straight up the middle of the compressor map.

You may or may not find the HX30 compressor to be a restriction at higher rpm. It's too close to call A flap valve to bypass it (with a wastegate actuator set to ~35psi triggered off the HE451) would cure that.

 

[Dougal]

Thanks Dougal. I have been going back and forth on the big turbo selection question.

Eventually I decided to just bolt it up and see how it goes. A guy stamsas in Norway has hx35/hx52 compound on a 606 and there is a dyno run on facebook which has both of them spooling quite late. Another guy Hans Persson in Sweden run straight hx52 but very late spool. All hx52 have 16cm housing, but larger turbine wheel than he451. So I think the smaller wheel on mine with similar housing cm should be about right.

I had not considered the flap compressor bypass until seeing blacksmoke racing on facebook with he221w under hx50 sequential. It has a gizmo like that. Somehow. Does the hx30 super contribute anything at 4000-6000? It is still presumably cranking away at 40psi or whatever. I did get the super so that should help.

Anyway now just have to either build the manifold or get someone working on it.

Cheers,
Karl

In my calcs I've got the HX30 fully wastegated at higher rpm to keep boost down. Otherwise you'll generate too much boost with them compounding and too much boost robs power. So once a HX30 is fully wastegated it should just be idling around and the compressor geometry that works very well when compressing air becomes a terrible flow path for air that doesn't need compressed.

Whether it is a significant restriction or not I don't know. Simply checking boost gauges on either side of it will let you know once it's up and running. But it's also only an issue in the 4000-6000rpm range which depending on the use of the car you may not even use much. It's going to have plenty of power either way. Power is showing pretty flat from 4500-6000rpm.

I would expect the HX35/HX52 to spool late not due entirely to the HX35. You need a small enough small turbine to create boost early and produce the extra airflow through the engine needed to get the big turbo up and out of bed.

 

[Atg]

Sorry this was a double post of the above. My browser did not update. I edited it down.

Stamsas in Norway is "coal smoke racing" on facebook in case anyone is interested.

Hoping not to melt the wheel in the hx30 super. I will run a d200 logger to capture egt by cylinder and maybe afr as well as iat, boost, and emp. Maybe run two intercoolers to save the little

So whether the small turbine maintains a pressure differential with the wastegate open depends mostly on the size of the wastegate. Going higher fuel than 120cc would then require a smaller wastegate if it requires compounding to get enough air.

I saw dieselmeken ab recent runs on two om606 engine dyno on facebook. They got about 490 crank hp on the stock manifold with 140cc and a big turbo at 35psi boost. They got about 550 crank on a twin turbo with 180cc minimum so on an 8mm pump at 50psi. Stock cams. No egts were plotted. 180cc appears far too much fuel to do anything but burn out for a bit on 50psi. The twin turbo motor had more down low but still not fully spooled before 3500rpm. Check it out on the videos. He posted photos of the data sheets.

I have seen Alcaid talk about an audi valve but I think it is for blow off. I do not know of a good diverted valve for the compressor bypass. I will study blacksmoke racing's sequential setup. I think their wastegate actuator diverts a flap in the exhaust stream away from the small turbine. They have both turbos plumbed straight to the intercooler I think, or a flap which bypasses the small compressor by opening when the large turbo begins to make more boost than the small. That flap is inside the intercooler in their build, and has no actuator. It just seals when pressure in intercooler exceeds pressure in small turbo boost tube I think. Their exhaust diverter flap was pretty bent by exhaust when they pulled it apart after last race though. Why they used that instead of a wastegate I don't know. Maybe it provides a smoother pressure transition.

A bit to figure out. I plan to use a synapse engineering wastegate 50mm, but could just use tial x2 big and small, and control them both with solenoids and the d200. Or use d200 to control the compressor diverted if needed.

Cheers,
Atg

So I've got a few spare minutes to run some calcs. So is the intent to run the 16cm 451 ungated or with an external gate?

I'll do a run with the 451-16cm ungated at 120cc and see how it looks.

*edit*
So I've run some ungated calcs and this is looking surprisingly good. You'll need a huge wastegate to bypass the HX30 pretty much completely from 4000rpm to 6000rpm.
Power should be around 350kw (crank).
Drive pressure is below boost from 4000-6000rpm.
The 451 runs straight up the middle of the compressor map.

You may or may not find the HX30 compressor to be a restriction at higher rpm. It's too close to call A flap valve to bypass it (with a wastegate actuator set to ~35psi triggered off the HE451) would cure that.