Thread: Another 120 project begins

Speaking of the pump specifications, I've scanned the pump test data & parts list for the 4BD1-T pump in its entirety and uploaded it to Files / Technical Manuals section of AULRO (link below).

http://www.aulro.com/afvb/local_links.php?catid=7

I have a hardcopy of the same document for a 4BD2-T pump, and can scan and upload if anyone is interested.

Cheers

Bojan

Originally Posted by Dougal

We're in a situation where any data is going to help. It's not difficult to use the injector pipes and injectors from the same engine. Why bother with a calibrated reference when you've got the real parts?

Recently on 4BTswaps.com one of the members had has injection pump tested for delivery both with and without injectors attached.

Results here: Got my VE flow benched! results inside!!!
and here: Got my VE flow benched! results inside!!! - Page 2

Cummins VE pumps can do 190cc at 3000rpm with an aftermarket fuel pin and differeng governor springs. But our engines flow air better.

I agree on using the stock injectors and lines, I just wanted to point out what is done in practice, to calibrate these pumps.

I've only had time to read a few posts from those threads you linked to, but one of your posts caught my eye from the 2nd link.

Honestly I can't see larger injectors being any help. Your fuel flow is limited by the fixed volume pump, not the injector.

I used to think the same, and had trouble reconciling why they claim HP increase with larger injectors. That was before reading Taylor's 2 volume book The internal combustion engine in theory and practice (not sure if I have the title 100% correct).

This book contains a lot of data with photos from laboratory experiments. And some show clearly what happens when injector size is changed to optimum for the fuel rate injected.

At large rates with small injectors, the finish of injection duration at the injector lags by a considerable number of degrees of crank angle behind the injection pump. This hurts BMEP and torque. Clearly when the fuel rate is increased significantly above stock, not only is the finish of injection at the pump increased (with stock plungers) but the injection into the combustion chamber is further increased past TDC.

Also at high PR and thus air density in the combustion chamber during injection, the spray from larger nozzle holes penetrates further and combustion is improved.

The reason I stated that larger injectors won't help is to crack these guys out of the petrol engine tuning. Where fuel is a constant pressure with plenty of it and the injector is the bottleneck.
I can appreciate fully the effects you mention from Taylors book (which I haven't read), but trying not to confuse the issue too much at the same time.

The aftermarket larger cummins injectors just seem to be good at making smoke. Possibly a mismatch between injector spray angles and the piston bowl.

Don't have too much progress to report but am finally about to get some garage time to put the engine together. I've cleaned the block and head mounting surfaces - gave it a clean with diesel and a Stanley blade to remove light surface rust, followed by a thorough clean with paint thinners to remove any oily residue.

Replacement 4BD1T headbolts have been ordered. Thanks for the part numbers John! Isuzu were great once they had the part numbers - the early 4BD1T bolts were replaced by the later ones, as you suspected.

Originally Posted by Bush65

4BD1
509009-0180 (14 required)
509009-0170 (4 required)

Early 4BD1T
894150-0250 (12 required)
894150-0230 (2 required)
894150-0240 (4 required)

Later 4BD1T
894367-4380 (12 required)
894367-4360 (2 required)
894367-4370 (4 required)

I presume the later 4BD1T bolts are improved material or heat treatment, not different size, to overcome some issue found earlier.

Had a chat with Isuzu about what type of molybdenum disulfide grease to use - they use a Fuchs Lithium based high temperature grease with MoS in their service centre. I ordered a small tube and am pretty sure its Fuchs RENOLIT MOREPLEX 2 HV, but will confirm once the order arrives.

In other news, I've got my hands on a copy of "Turbocharging the internal combustion engine" by Watson & Janota, and am soldiering through to wrap my head around all things turbo.

It quickly became apparent that my assumption of 60% intercooler efficiency may not be particularly accurate, and that it has an impact on sizing calculations. I have a reasonably large front mount air to air intercooler, and comparing published IC effectiveness figures for similar sized FMICs, it's more likely to be in the mid 70% range at highway speeds. The intercooler has a core area of 550mm x 235mm x 65mm and has virtually no obstruction to airflow. Fortunately an efficiency increase works well with the T25 compressor. I've plotted the 75% IC effectiveness points corresponding to engine power in light colours next to the 60% points I had earlier:



Increasing IC effectiveness makes the compressor efficiency peak lower in the RPM range (1900 to 2300, depending on load), which should provide more torque backup as well as increase engine efficiency at highway speeds. It also puts my 150kW / 200hp point at 3000 RPM on the compressor map.

Reading about turbine design however has only confused me further... or rather made me realise that what I thought I knew is wrong.

I understand that turbine housing A/R ratio is an important descriptor for a vaneless housing. The A/R ration determines the gas exit angle from the housing (volute) onto the turbine wheel. As I understand it, the angle is almost independent of mass flow rate through the turbine.

In a vaned turbine housing however, this angle is determined by the geometry of the nozzles, and the only purpose the volute serves is to deliver a uniform flow to the nozzles. In otherwords, the A/R ratio of the VNT / VGT turbine housing has little impact on turbine performance / choke flow rate as this is essentially determined by the nozzle geometry. So it depends on the range of nozzle movement possible which, of course, isn't published!
The nozzles have a considerable range of movement (see below). The graph on the right shows the influence on choke flow that a 20 degree change in nozzle angle can have (80, 70 and 60 degrees from radial shown).


VNT ring out of the housing


The good news as far as I can determine are that:

1.) Fixed nozzle turbines typically achieve a higher efficiency than nozzleless ones, but over a narrower flow range. Variable nozzle geometry should theoretically extend the range of flow where high efficiency is achieved.
2.) All other things being equal, a high trim turbine wheel flows better than a low trim one, so there's hope for the VNT-25 yet.

I think it will be a case of throw it on and see what happens. I may have to drill a pressure tube and a thermocouple housing both pre and post turbo and do a few test runs. But first I need to put it together and into the truck. Any ideas for a T3 to TK100 (early Chrysler T3 turbo mount) adapter? From an airflow perspective, I understand it is better to gradually blend the larger T3 port into the smaller TK100 one, rather than having a sharp step between the two - can anyone confirm?

Cheers

Bojan

I've found some interesting stuff about VNT turbos recently too.

The A/R on the housing is pretty much the max the vanes open to. The adjustment range is all on the side to mimick a smaller A/R, not bigger. Obviously you get the best efficiency when the blades aren't changing the flow angle and this is best kept for the high flow and high rpm to get max power from the engine.

I've also found a very good way to correlate max corrected choke flow based on exducer and A/R. So what are those measurements for your VNT25 again?

The learnings I've done over the last few weeks mean my GT2256V will be up for sale. It's just no good on an engine like this. It could be modified with the higher trim (larger exducer) GT2260V turbine wheel. But it'll still be ****ed all over by a wastegated TD04HL-19T.

As for intercooling. I'd hope you can do better than 60%. But I use numbers on the safe side for all my calcs. I'd rather have the machine beat predictions than not meet expectations.

The variable nozzle turbine housing has an 0.64 A/R housing with a 52.7mm inducer and 46.2mm exducer (77 trim).

Originally Posted by Dougal

The A/R on the housing is pretty much the max the vanes open to. The adjustment range is all on the side to mimick a smaller A/R, not bigger. Obviously you get the best efficiency when the blades aren't changing the flow angle and this is best kept for the high flow and high rpm to get max power from the engine.

Are you sure this is the case? This is what I used to think until I started reading the radial flow turbine section of Watson and Janota... now I'm not so sure.

To summarise my understanding of turbine efficiency / choke flow...



As I understand it, the nozzle angle affects the throat size, which is one of the factors determining the choked mass flow, but it also determines the gas flow inlet angle, which affects the turbine efficiency for a particular mass flow.

On the diagram showing the rotor blades and a single nozzle (below left), resolving the gas flow inlet velocity (C4 on the vector diagram on the right) into a tangential component (Ctheta4 on the diagram, where "theta" looks like a B - apologies for the poor scan) and radial component (Cr4) relative to the rotor blades, the tangential component of velocity (together with mass flow and the effective radius of the rotor blades) determines torque applied to the shaft. Rate of energy transfer at the wheel then becomes the product of torque and angular velocity of the rotor. The radial component of the inlet velocity determines how quickly the gas moves through the rotor (i.e. it impacts on the choked mass flow).



At very low mass flows, the turbine is most efficient when the flow is almost tangential to the rotor blades so that most of its energy is transferred into the wheel. Because the mass flow is low, the radial component of velocity doesn't need to be very high to clear the gas through the rotor blades.

As the mass flow rate increases, the radial component of inlet velocity needs to increase for optimum efficiency in order to prevent choking through the rotor blades. Torque applied to the wheel increases even though the tangential component of velocity decreases due to the greater mass flow through the rotor. So as mass flow through the turbine increases, the inlet angle at which the greatest efficiency is achieved shifts from tangential to radial direction, in order to allow (or rather force) more gas to flow through the turbine without choking.

Frictional losses occur due to nozzle angles not aligning with the gas inlet angle dictated by the A/R of the housing, but these are considerably lower than the gains in turbine efficiency by matching the gas inlet angle to the mass flow through the turbine. In other words, the nozzles reduce the overall efficiency of the turbine, but extend the range of mass flows at which the turbine operates at optimum efficiency.

On a vaneless / nozzleless turbine, the angle at which the gas enters the rotor is determined by the relative area of the turbine scroll (A/R ratio of the housing), and the angle is almost independent of mass flow through the turbine.

It is now my understanding that the A/R ratio of the housing on a VNT turbine (or a fixed nozzle turbine for that matter) doesn't have a large impact on the choke flow through the turbine. Obviously, the exducer diameter and the combined throat area around the nozzle ring also contribute to the choke mass flow, but at which flow rate they start becoming dominant is anyone's guess.

Hence my confusion... and belief there's no way of predicting what the turbine choke flow will be without trying it out.

Cheers

Bojan

I pretty much agree with all that. The eureka moment came when I was studying the only two official garrett VNT turbine maps I've been able to find. Max choke flow on each was comparable to the same exducer size and housing A/R on a fixed geometry housing. Vanes at full open and it's behaving just like the A/R on the side and the wheel inside say it should. The vanes are just a slight efficiency killer in that position.

Remember the blades on an angle to modify the flow are an inefficiency. To maximise the energy extracted by the turbine and power of the engine you need to combine highest flow with best efficiency. This is when the blades aren't changing the flow direction.

Choke flow prediction, I'm pretty sure I've got it. I've got a table with all the available turbine maps in sizes that interest us with A/R and exducer size. Suffice to say I can empirically line them all up with choke flow based off A/R and exducer.
There is a drift in the absolute prediction numbers as turbines get much larger and rpm gets lower. But the trend is very clear and interpolating within the data I have is quite straight-forward.

It makes sense for the manufacturer to use the equivalent A/R ratio to specify the choke flow for the turbo. What choke flow do you have for a housing that has an 0.64 A/R, a 52.7mm inducer and a 46.2mm exducer (77 trim)?