Large Turbine + Tiny A/r Vs Small Turbine + Small A/r

[Skepticism]

This has been nagging at me a bit. Apparently Garrett's created a T3, twinscroll 0.61 A/R housing that would presumably go on something like a GTX3067. It seems faaairly undersized (understatement) and makes me wonder what Garrett had in mind.

What would its characteristics be like in comparison to a T28 turbine with the .64 a/r housing? Twin scroll aside. I'm not sure how to calculate which housing is actually bigger given the difference of T2/T3 and single/twin scroll, but my guesstimate is that the .61 is actually smaller.

Edited December 17, 2014 by Skepticism

[zebra]

You cant compare A/R Ratios of different frame size turbos

[discopotato03]

A few things to consider

Firstly the turbo (turbine housing) has to fit your manifold/engine so John West any that don't .

Secondly Garrett do use the same turbine housing castings at times for a few different sized turbines ie up until recently GT30 and GT35 turbines in differently profiled common housings .

Mostly GT28 turbine sized turbos don't bolt up to RB manifolds unless they are unique HKS hybrid housings with T3 flanges , Garrett pretty much used T2 flanges on GT28 turbos/housings .

Garrett is obviously making a range of turbine housing ARs to give users a bit of flexibility with different capacity engines . For example if you had a 13-1600cc engine and you wanted to use a TS housing then 0.61AR probably isn't unreasonable . You'd think its small for a 2-2.5L engines but even so on a low revving diseasel application maybe fine . For petrol a bit small IMO .

Most people agree that comparing single and twin scroll ARs is a head flick , the whole point of twin scrolls is to halve the number of exhaust "putts" into each side to make the engine scavange better and reduce exhaust reversion .

I believe that when the engines loaded up and high in its rev range the total flow capacity of the turbine/housing combination is more important and the overlapping exhaust pulses get so close together that the TS advantage tapers off .

In the heavy duty engine world , most big automotive diesels , the housings are sized to give the best performance and efficiencies under full load because that's where they spend much of their time .

Different with a road car engine because you mostly can't use its performance potential for more than short bursts and stay legal speed wise . In some ways having a wide rev range shows up the limitations of centrifugal turbochargers because its difficult to make them work effectively across the whole speed range of your average petrol engine .

Twin scroll turbocharging adds another dimension of complexity (manifold/turbine housing/wastegating) but done properly can help widen the forced engines power range over single scroll setups . I believe done properly it achieves this by improving the engines performance off boost and allows better efficiencies up high if the turbine/housing/gates are sized to suit the users top end wants . Everybody's wants are different so each has to define from the outset close to what they want to achieve .

Personally I think GT28 gear is a tad limiting for any single turbo RB except maybe a mild RB20 , twins and 26s are a different story . I think you can do better for an RB25 and we'll soon know if the 0.63AR single scroll housing on a GTX3067R is responsive enough for even a conservative RB25 .

I have read that Garrett wants to do larger AR housings for GT28 turbos but I reckon there are limits to how high a capacity compressor a GT28 turbine can drive regardless of how big the 28 housing gets . People reckon the 56T 71mm GT compressor was getting laggy with the 28 turbine and I don't think the situation gets much if any better with the GTX67 or GTX71mm compressors .

Borg Warner has shown that turbines can get larger in relation to compressors if they have less mass/inertia but that means using expensive exotic materials and Garrett doesn't seem to be interested in going there .

Hope this helps cheers A .

[M@&k]

Going off topic; but 0.63A/R GTX3067 results are in. Be interested in your thoughts there disco.

[Skepticism]

Thanks for the answer disco

Bear in mind that I've got an SR20 - I just thought SAU would be the more interesting place to post this as I've seen a lot of in depth discussion on turbos

Upsizing the turbine wheel seems interesting for SRs though, as going for the large T28 housings is a generally a waste of time. Maybe because unlike RBs, SRs make most of their torque in the mid range and fall off quickly, so need a bunch of timing up top to make power - so maybe the SR head limits things and breathing isn't such a huge priority. With bigger housings there's still a power increase, but with substantially worse lag and boost threshhold. Conventional wisdom is that a bigger comp wheel on the same housing will yield similar power with better transient results. Nobody has ever really tested a really small A/R with a bigger wheel before though. I have no idea what would happen. I guess that's my main question.

Edited December 17, 2014 by Skepticism

[discopotato03]

I haven't yet seen the 63 3067R results so busy ATM .

All I know about SR heads these days is low port high port - and a little about the GTiR variation .

What I do know is that sometimes using a lower max flow compressor for a given turbine size can give good results especially with low to mid range power in mind . I believe it comes back to having equal pressures across the cylinder head (inlet and exhaust) when running in the most useful part of the rev range .

Gotta run cheers A .

[discopotato03]

Some examples of lower capacity compressors were done by Mitsubishi but in some model Garretts you have a choice of trim sizes , you don't always have to go for the highest capacity one . In their later models you don't get too much say but there is a large range with small jumps for the GT28 series using GT or GTX compressors .

Slowly more pics and details are filtering out on the GTW turbos and I have a bad feeling that they are larger BB and plain bearing cartridges going off the pics . The larger centre section looks like the one used in the BB T04/60-1 and T04Z units . The turbine dimensions look like (from memory) T04 in N o and P trims and the smaller ones some other T series offering .

The compressor backplate looks like machined aluminium not the ring type adapters used in GT25 based BB cartridges ie GT2554R - to GT3582R .

The 9 bladed billet compressors are a step in the right direction and lets hope they make similar things available for the GT30 and GT35 based BB turbos .

Where abouts is the results of the 0.63 AR GTX3067R ?

A .

[M@&k]

Where abouts is the results of the 0.63 AR GTX3067R ?

In the "Looks like Garrett slipped another GTX30 variant in" thread. Also in RB25 dyno results thread.

[hypergear]

read through my thread. Its got all combinations mentioned above with results.