Torque - Discussion, Physics, Displacement And Ideas

[GTRNUR]

the fact that you could breath on the throttle and the thing would hit 10k rpm in a split second made it the quicker car.

Thats where confusing horsepower combined with gear ratios producing a more usable torque to the rear axel can be confused with torque at the engine. Simular to the mines 34's that run a getrag with tiny diff gears yet rev to 9K+. They are so responsive and quick because of the driveline makes better use of the high RPM horsepower (torque at high revs).

In a street car having monster amounts of torque from low in the RPM range is what makes the car more enjoyable to drive. As Paul suggested, a torque comparison between a RB20 and 4g63 (in naturally aspirated form) for all given RPM points would be a good comparison. Comparing naturally aspirated and non-vct engines limits the comparison to just geometry.

4g63 sohc - 85mm bore x 88m stroke. 5.9" rod and 1.7:1 rod ratio, 8.5:1cr - 157nm @ 4750rpm

rb20e sohc - 78mm bore x 69.7 stroke, 122mm rod and 1.75 rod ratio, 9.5:1 - 171 nm @ 4400 rpm

Interesting... nearly the same bar the compression which has me thinking the 4g63 is the better engine geometry wise.

[bradsm87]

rod length to stroke ratio as well as bore to stroke ratio play a much bigger role in torque that capacity per cylinder imo.

[GTScotT]

Ive heard evo's run reallllly high psi from factory...evo X in particular..is that true?

This is really off topic, but, evo's 4 - 9 run 1 bar from stock. Unsure of the 10.

[GTRNUR]

IMO the main ingredient in the mix here is the bore x stroke and the turbo in question. A square engine is more than the best of both worlds, its almost the best of both worlds. Unmatched torque and an ability to rev. The SR is square: 86x86, where as the 4G has a minor offset: 87x85. That slight change put together with a really well flowing head from the factory results in a motor that wants to rev infinitely and has loads of power and torque.

Surprisingly though, the EJ25 in question is nothing great in that sense, its something like 75 x 99. BUT, dare I say the whole 'boxer' principal plays a major part.

What you left out there is that the Sr runs a short 5.35" rod and the 4g63 uses a 5.9" rod. The 1mm differences in bore and stroke alone can not take full responsibility for the different charactoristics of the engine. No single charactoristic is responsible for how an engine behaves. The combination of cam configuration with geometry and CR will also make a huge difference. Anyone care to track down the camshaft data on a SR20det and 4g63?

The 'boxer" principal? In the end that has little bearing on anything. The main benefit of a flat engine is that it is shorter, same as a V engine. It also can allow an engine to run a much larger bore as the cylinder spacing is larger for the layout. Without meaning to over simplify... in each cylinder the pistons and rods still go up and down from the center line of the crank, the same as an inline engine.

An interesting thing to note though is that the EJ engines tend to have massive rod journes in comparison to RB engines... I dont know about 4g63's though. This may be in part be an effort to reduce crank flex as the all alloy engine of an EJ in standard form runs pretty tight bearing clearences from what I can remember. (sorry... getting off topic here).

[SoFreshSoClean]

http://www.autospeed.com.au/cms/A_2807/article.html

Here is a better way of comparing engine designs

[GTRPowa]

I agree with jarrod. You'd have to compare the actual torque curves.

My argument is that the more max power, the less you will "feel" the torque. Although the torque is there, is doesn't compare to the sudden woosh shortly after.

So for arguments sake, if you had a 300awkw GTR next to a 180awkw WRX, the difference between the torque and power on the GTR would make it 'feel' like it has less torque.

However with the WRX since the delivery is more linear, it feels like there is more torque.

[wolverine]

and high flowing is IMO no form of a fix. Run either a GTRS

I have run both (back to back) and there is a bees dick in it.

(GCG vs HKS GT-RS)

[Gerg_R31]

if this is a ej25 vs rb25 could someone fill in the following

ej25

bore

stroke

compression ratio

induction flow in cfm

turbo flow

turbo pressure ratio

boost

turbo surge flow

cam lift and duration

rb25 neo

induction flow in cfm

turbo flow

turbo pressure ratio

boost

turbo surge flow

and i can use my dyno program thing to put up a couple graphs that show the torque and power curve

[VB-]

i like the idea of the thread, but as others have mentioned, i think other things effect torque alot more than cc per cylender

so eliminating driveline and head differences (where most of the difference is IMHO), lets take two motors, of the same capacity. one is short stroke/big bore, one is long stroke, small bore. one should make more top end power, one more low end torque. i think most will agree on this.

i am curious to know though, for the purpose of this discusion, take two motors, same head design etc, only difference being bottom end, one is a 2L 4 banger, one is a 2L inline 6, both are square motors (Bore:stroke ratio), what will the difference in torque curves be?

[GTScotT]

I have run both (back to back) and there is a bees dick in it.

(GCG vs HKS GT-RS)

I have read it many a time, and have faith in the GCG product, yet there are numbers and there are numbers.

HKS themselves have also produced a document for the new 2835kai. It shows the GTRS vs 2835 and 2835kai, there would be a bee's dick between all of them, yet compare the performance of a 2835 equipped skyline to that of a GTRS or highflowed. Somewhat different to say the least.

Power isn't everything, but delivery is king.

End of the day, delivery is what we are talking about here, more so than actual torque figures. Food for thought, and please lets not turn this into a turbo battle.

[Kwyjibo]

http://www.autospeed.com.au/cms/A_2807/article.html

Here is a better way of comparing engine designs

Surely one of the best ways of comparing engine designs is by comparing the inputs vs the outputs. I.e the most efficient engine should be the one generating the most power with the least fuel and air being put into it.

[Nee-san]

Surely one of the best ways of comparing engine designs is by comparing the inputs vs the outputs. I.e the most efficient engine should be the one generating the most power with the least fuel and air being put into it.

Correct.

[zebra]

Does anyone have a copy of an engine designing program...i dont have mine anymore, would be good to punch in some numbers and test it all in theory.

[VB-]

Surely one of the best ways of comparing engine designs is by comparing the inputs vs the outputs. I.e the most efficient engine should be the one generating the most power with the least fuel and air being put into it.

fuel you can measrue, and comparing it to output is called BSFC, brake specific fuel consumption. its shithouse for comparing cars because all the numbers are very similar

measuring air and comparing against output i have never seen, dont know what you could use to accurately measure such a large volume of air. you could calculate it providing you had an accurate VE number?

zebra, download a trial version of engine analyzer. i think you can find cracks for em too

Edited July 25, 2010 by VB-

[SoFreshSoClean]

Surely one of the best ways of comparing engine designs is by comparing the inputs vs the outputs. I.e the most efficient engine should be the one generating the most power with the least fuel and air being put into it.

We are talking about power and torque not efficiency remember..

[discopotato03]

Hmm , good thread this one and I have a Subie an R33 and an Evo 6 .

The real answer is that there isn't one because there are so many variables .

The most significant one is that its impossible to directly compare a four and a six cylinder , even of the same capacity , because the firing orders are different and so is the number of crank degrees per power stroke .

The given example was a 2L four and a 2L six , the fours power strokes in theory create more shove but the sixes ones are closer together ie 120 degrees for the six and 180 degrees for the four .

I disagree about the square or same bore and stroke , the best four I ever tried was Nissans old FJ20 and they were 89mm bore + 80mm stroke . From memory the rod centres length is 140mm . They make great torque and rev like a turbine if the exhaust system isn't restrictive .

The point you have to remember is that the piston crown area is that which combustion pressure acts upon and the large bore size is that which allows you to have a lot of valve area before they hit to sides of the bores .

Anyone who tried to hot up an RB20 soon worked out that you can't really increase the valve sizes because they clout the block .

Something interesting four pot wise was what Nissan did to the "SR20's" when they used them in the 2L Super Tourer class . The rules allowed them to do really extreme things with the blocks and heads and what they ended up with was roughly the FJ20 bore and stroke and possibly more weld material than head casting . I thing the engine ran in the opposite direction and they turned the head around on the block drove the cams from the opposite to OE end of the head .

Turbo rally engines are very different to tarmac race in that they have to make killer torque from low revs and are strangled by turbo restrictors so swap cogs no kidding at 5-5500 revs .

The 4G63Ts are not a square engine and like many production transverse engines need to be short to fit with a transmission across the engine bay .

They make good torque because of the rally homologation development and are in a reasonably high state of tune for a production engine . Things like the CR being around 8.75 - 8.8 helps make them crisp and make more low end torque than say an SR20 at 8.5:1 . The largish intercooler in the nose of an Evolution Lancer means you can have the higher CR and higher than average boost pressure without the rattle rattle dramas . They don't get short changed with radiators and many have a proper air to air oil cooler as well . People like Corky always said that thermal issues are the achilles heel of forced induced engines so if you can beet the extra heat they generate they become a lot more reliable and effective particularly in higher states of tune .

The Turbochargers . Subaru initially used IHI turbos on their EJ engines eg the VF12 on the AMD Liberty RS Turbos . When Rex turned up it got a Mitsubishi Heavy Industries (MHI) TD05 dryer and in the later 90s a TD04 because everyone bitched about turbo lag . What they could have used as a better compromise was the large turbine version of the TD04 ie TD04H instead of TD04L , actually the rare auto wagon JDM early WRX did get this one std .

My opinion is that Fuji botched the 90s era EJ20 turbo engines exhaust header wise , a real dogs breakfast except for the later factory twin scroll header manifold and turbo system - TS turbine housings . BTW they had both IHI and Mitsy TS turbos too , search TD04HLA-19T ...

EJ20 turbo engines with the stone age header system (single scroll system) get real laggy real quickly with larger turbos .

Its partly because flat fours have a screwy firing order to stop the engine lunging from side to side with each power stroke . They fire the cylinders in each bank then swap to the other side and do the same again . Looking from the front ist left front left rear then right front right rear , the firing order is 1324 rather than an inline fours 1342 .

So you say ? Problem is that you get two exhaust events from each side close together the a longer pause before reverting to the other side . Just to screw things up the nearside head has a long engine pipe around the front of the sump and the turbo side one is real short before both joint and feed the turbo via the up pipe . The result is unevenly spaced exhaust pulses and scavenging , its also the reason you get that silly boxer burble when you use a storm water drain for an exhaust pipe . The twin scroll EJ20's sound more like an inline four and are more of a basic fix than an upgrade .

Just on boxer con rods , yes the big ends are big and narrow because flat engine cranks are quite short so the rods/webs etc need to be wide and skinny .

And yes the split aluminium engine cases can have rigidity problems at times .

Mitsubishi got wise in the Evo 4 era and decided to go with a twin scroll turbo system and it works really well . Size for size the Mitsy turbos tend to work better in Evo guise than Rex guise mainly because of the twin scroll turbine housing .

Those Evo 4 TS turbine housings were small (for a TS) at 9cm but Mitsy soon worked out that the larger 10.5cm ones worked even better ie virtually no more lag and less exhaust restriction .

Your typical Rex single scroll housing is usually in the 6-8cm size range and the larger ones can be a bit lazy .

Now to the RB6's . These are very much late 80's thinking (RB20/26) with the 25 being a less costly destroked version .

The R33 brought variable cam timing and the R34 got non hydraulic buckets and a bit more VCT sophistication .

Some like SK think that the valve size ratio is wrong and that the exhaust ones could have been a bit larger . I don't think I'd do an RB30 without doing both because any breathing restriction means less complete cylinder filling esp off boost and it cost you torque - and power up high . It would be interesting to know what Toyota does valve size wise with 2JZ GTEs .

Good grunty low to mid range performance in production cars comes from making good part throttle torque without a lot of revs . A critical part also is gearing because if its not well chosen to make use of this part throttle torque then its largely wasted . With turbo engines the secret is having a high 8's CR and a head/s that breath really well . VCT is a way to have healthy cams and opening them up to have less overlap and high trapping efficiency at low to medium revs . Good intercooling is a must because higher CRs + healthy boost means heat and if not dealt with detonation .

Multiple throttles are a nice if expensive option because they offer the least throttling restriction and while shut reduce the reversion effects of healthy cams .

All Wheel Drive - the permanent variety . No real substitute for having 100% of the weight of a car on its driving wheels . Double the contact patch area of FrWD or RWD's and they share the drive loads . Two of my cars are permanent AWD and that puts an end to the spin tramp my R33 gets if I boot it off the line in the dry or over every painted line on the road in the wet . The torque you can put to the deck with them kills the R33 off the line and they don't feel taily at all .

Power delivery wise the single turbo RB25 is crying out for a twin scroll turbo because it can't scavenge or breath out as well with the SS turbo or get the turbo spinning up as easily either . It may be a legacy of the fact that RB thinking came out of the 80's when twin scroll twin integral gate turbos virtually didn't exist .

The 4G's are a reasonably well designed 4 cylinder twin cam engine but there no F1 engine . The overall package functions well and everything is designed to work together . They went through many revisions close together because Mitsy wanted to stay competitive in Group A Rallying and competition R & D soon shows what works and what doesn't .

I don't think Nissan went to anything like as much trouble with the RB family and even the 26's had a short life in the competition world . Nissan could have done the same things that Mitsubishi did but I'm told they were in financial hot water and the RBs and SRs weren't going to get beyond the 90s anyway .

Just for the cc list Porsche had a 3L four cylinder for a while , the largest petrol four I know of .

Paul go for a ride in a mildly modded Evo 6-9 and see what you think , more like a toned down go cart than a four door sedan but its that too BTW .

No fingers left , cheers A .

[D_Stirls]

I was going to mention that volumetric efficiency has a large part to play engine torque characteristics of an engine.

[DVS JEZ]

Wow. Nice write up. Must have taken ages

[discopotato03]

Volumetric efficiency and dynamic compression ratio are directly related meaning if an engine can breathe well enough to completely fill its cylinders the dynamic CR will be close to the measured CR .

The other side of volumetric efficiency is the hot side or how well and completely the cylinders blow down whilst the exhaust valves are open .

There is a basic problem with exhaust driven air pumps and that is attempting to keep the exhaust and inlet manifold pressures as close together as possible . When you can achieve it you get the ideal pressure balance across the head meaning it has the best chance of scavenging or blowing out the residual spent gasses and heat at the end of the exhaust cycle .

The basic single scroll turbo system can do it with very large AR ratio turbine housings but don't expect the thing to start pumping air too early .

Also there are two big advantages twin scroll systems have over the single one . Firstly the engine has less resistance to venting its cylinders and scavenging properly AND the pressure rise across each scroll is greater purely because it was lower to start with . The rise in velocity of the exhaust gas is greatest just as the exhaust valves crack open and its this sudden bust of hot gas that gets the turbine accelerating sooner or lower in the engines rev range .

In the diesel world TS turbocharging was designed to help the engine make strong torque at lower engine revs without losing anything further up a diesels generally short/narrow rev range . It can at times be a way of getting larger capacity engines low down torque without the capacity increase .

Larger cylinders are usually hard to beat but getting all 43 beans out of any engine is better than only getting 35 out of a slightly larger one .

Capacity increases . Most production engines these days don't have real thick cylinder walls mainly because more material costs more money and adds avoidable weight to increasingly heavier cars .

The blocks bore centers can be closer together meaning a shorter more compact engine , manufacturers like Toyota/Nissan/Mitsubishi have four cylinder lumps used in transverse as well as inline apps . Examples are Toyotas 4AGE , Nissans SR20/Mitsys $g family . Short engines are easier thrown in sideways between the towers of FrWD and some AWD cars . I pity people who drive east west FrWD V6 cars ...

Anyway you can increase the capacity of these fours by increasing the cranks stroke and it doesn't work out too badly if say the blocks height is raised so that the con rod length is increased - better rod/stroke ratio . Some would argue that stroking is a bad idea but remember that these are production engines not race ones and Joe average isn't going to rev it past 4000 most of the time .

Example engines are Nissans RB30 , Toyotas 7A , Mitsys 4G64 . Nissans stroke is still less than the bore size but then again the smallest "big bore" RB engine (RB25) is very much a short stroke engine for its bore size .

No one in their right mind would increase stroke length and decrease bore size for a given production engine to argue that stroking is a bad idea .

The increased cylinder capacity will always result in more air and fuel to compress and burn and provided the engine is not revved past its safe mechanical limits it should make more torque than a std stroke version .

There's a lot more to it than that because the manufacturer designs an engines breathing characteristics to promote torque in a given engine speed range and when the cylinders are larger the gas gas speed through everything is greater for the same revs . Also the gearing is designed to match what the manufacturer wanted from their std product and its state of tune . So IMO things like an RB30DET in an R33 should have a taller diff ratio , thats what I think Nnissan would have done had they made such a car .

Other places to be , cheers A .

[simpletool]

Any comparison with mid-range torque for boosted engines needs to take into account FIRSTLY the boost pressure. Have a look at any latter model Evo (especially the 10) and notice how it runs big boost in the mid-range and tapers massively at the top end. An Evo 9 runs over 1 bar in the mid-range and an Evo 10 over 20psi, but this falls off massively as the revs rise (this is why EVOs make such big gains with flash tuning - to level out the boost curve and lean mixtures). This is the opposite of the R33 RB25DET, where stock boost rises at higher revs (presumably to make the R33 smoother, less snappy and more GT-like). WRX which also runs a small factory turbo with high-pressure which will make big mid-range torque but run out of flow up high, plus the heads on pleb-WRX are good for mid-range not top-end.

I'd say this difference in mid-range torque has a lot more to do with turbo size, variable boost pressure (in the late Evo especially) and head design than cylinder dimensions, rod-length or number of cylinders. The over-square dimensions of the RB-series (ideal for the RB26) coupled with the small turbo but low mid-range pressure on R33 RB25 was the worst of both worlds however - but this just means it's easier to get big gains. The factory's loss is our gain.

Edited July 26, 2010 by simpletool