Power and torque curve dips 2

[colinmseries]

I've just taken delivery of a modifyied i4 Duratech engine that's been taken out to 2.5lt The builder supplied a power and torque plot from the dyno and at 2200 and 4500 revs there is a pronounced dip of some 30-50lbs.ft I've asked him and he says this to some extent is always the case and is caused by resonance, would you agree?
I would add that I'm more than happy with his work and reputation but somewhat puzzled by this phenomenon.
Thanks, Colin.

 

[Marquis]

Hi, it's very engine type specific.
I have done extensive work on engine dynamometer and shadowed it with GT Power/Ricardo Wave 1 D cycle simulation codes for excellent correlation but more importantly until I got a "feel" for that type of engine. In my career I've worked mainly with Cruciform crank (conventional) V8s , 60 degree V6s, Boxer 6s and some straight sixes. There are definate trend with these types of engines almost regardless of make, such that when I see a V8 Torque curve with dips in the 1000 to 2500 rpm region I'll automatically look toward the exhaust system and then come toward the catalysts [this is assuming the engine does NOT have two stage variable length inlet manifolding].

Unfortunately my experience of in line fours is limited. I DO know that they are more sensitive to the inlet air box volumes and general geometries upstream of the plenum throttle, where as on the types of V8s and V6s I've worked on, quater wave resonators for NVH upstream of the throttle body made naff all difference to the measured torque curve.

A pronounced dip of the amount you say over such a big rev range sounds like alot more than just a "resonance" phenomenon- could be mismatched cam specs- to the inlet runner tuning lengths (lots of overlap with very short runner lengths?). I'm suspicious of the "resonance" explanation. Assuming it is a "resonance" effect, resonance can always be tuned out, if you have the freedom to redesign the inlet manifold/intake system- so I don't buy that it's "always the case". May be always the case if this engine is build in this kind of way.
Best most expedient/cost effective way to tackle/understand it would be to get someone to build a 1 D cycle simulation model of the engine. Even if the simulation doesn't correlate with the absolute measured numbers for the engine- I have enough confidence (and MORE confidence than a back street chasis dyno or inertia chasis dyno) that it would allow good understanding. Unfortunately, although cheaper than an extensive dyno test investigation would still not be cheap.

Good luck!

 

[SomptingGuy]

Adding to the above, sometimes dips in torque mysteriously appear exactly where the engine operates during a pass-by noise test. I wonder why. You never know, it might even be intentional.

 

[colinmseries]

Your comments regarding inlet lengths and cams ring a bell. The commission for the engine was to maximise low end torque and to give no particular mind to peak power. That was to be whatever it turned out.
During our conversations the builder expected to be using longer inlet lengths and said when they were tested, with the i4, he found they made very little difference. This surprised him as he specialises in 4 cylinder engines and had not seen this before.

 

[GregLocock]

I'd wander over to the LARC archives and read up on intake manifold tuning.

Incidentally, it may not be a coincidence that 4500=2*2200, for sufficiently large values of 2

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[bigpaul]

Almost certainly its primery pipe lenths 2.5 duratec do not
responed the same way as small engined He. Engine induction
lenth becomes more sensitive when correct but do not use same formula as small capacity

 

[PackardV8]

Greetings, Marquis,

Just the man I have been looking for! On many published torque curves for V8s I see in magazines and many of the torque curves I generate with the commercial DynoSim program, there is a noticable dip in the 2500-3200 RPM range. I have adjusted most of the variables in the DynoSim program without causing the dip to smooth out.

Most of these are performance engines, with larger porting and longer cam timing, but still street engines. Some of the dips remain even when switching from NA to supercharged, so it isn't primpary pipe length on the V8s. Where should I look for the causes?

thnx, jv.

 

[colinmseries]

Thanks for comment so far. I've searched LaRCNASA but didn't fnd much I recognised as helpful. Just to be sure we're are talking about the same engine: mine is the 4cylinder Mazda/Ford Duratech originally a 2.3lt and bored out here in England to 2.5lt (I wasn't sure from some of the abbreviations).
As I've mentioned, the builder expected to be using longer inlet lengths than usual but when he tested it on his dyno he reported that he couldn't get much difference. It's some way off being ready to use the car and see how it feels. I can't see how to upload the graph aand see for your self.
Colin

 

[Lou Scannon]

PackardV8, why should the dips disappear when switching from NA to supercharged?

 

[PackardV8]

A little theory and a bit of experience:

1. Supercharging changes everything about the inlet track. The same length manifold runners behave completely differently under pressure than they do under vacuum.
2. Compression ratios behave very differently when supercharged.
3. Observe the published torque curve for the turbocharged GM Ecotec four in the Saturn Sky. It is perfectly flat at 258 lb-ft from 2500-5300 RPMS. Not only no dips, anywhere, It goes nearly straight up, makes a near 90-degree turn and stays flat to the cut-off. Tne NA version is the expected arc.

thnx, jv.

 

[patprimmer]

Umm

I can see a little bit of experience, but where is the theory.



Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[Paulista]

I have read in a relatively recent reference-type of text book relating to modern developments in IC engines that the turbo-charged engine is a self-regulating (feed-back) type of device.

For this reason it is stated that engine torque vs rpm curves are more or less the same, say, between sea-level and higher altitudes. The same is definitely not true for a NA engine - power tapers off pretty dramatically specially at higher altitudes.

Not had much luck to actually model a turbo-charged engine as such (self-regulating), although tapping into waste power (energy) that would normally exit uneventfully in the exhaust and feeding back some of this power (energy) in the way of boost pressure does look like "power feed-back" .

So recalling some results from lectures and course-work (lab) on negative-feedback (NFB) as the latter applies to electronic amplifiers...... the curves of amplitude (power) response vs frequency get incredibly flatter when NFB is applied around the loop.

In adition, should the NFB loop include frequency-selective components, lo and behold you have bfore you a very fine tone-control !

Our lecturer used to emphasise "gain X bandwith product is constant", so if you emply NFB and reduce gain, the 3-db bandwidth goes up.

NFB also confers a great deal of immunity to extraneous factors such as for example power supply variations.

Could some - or all - or none (?)- of the above analogy be applicable to turbo-charged engines ?

 

[Lou Scannon]

"Published" torque curves are a little suspect anyway...
However, with an electronically control turbocharged engine, these days, boost pressure is regulated during maximum operator demand so as to deliver a specified torque curve, so it is no longer governed by volumetric and mechanical efficiency.

Paulista,
ruling out electronic control using algorithms designed to maintain constant charge density with varying ambient pressure (i.e. at maximum operator demand), I don't see that a mechanically wastegated turbocharger by itself fully compensates for reduced barometric pressure. All a conventional mechanical wastegate does is limit boost pressure to a calibrated delta above ambient, so a reduction in ambient pressure directly translates to a reduction in absolute boost pressure. I suppose you could have an aneroid (absolute reference) wastegate, but I've never heard of one. An even more noticable effect is seen in the spool-up portion of the torque curve, when the wastegate is still fully closed. This is a "bootstrap" operation for the turbo, and with reduced ambient density (whether due to pressure or temperature), free-floating (i.e. non-wastegated) boost pressure is drastically affected, due to the physics of turbocharger operation.

 

[Paulista]

hemi,
Agree that at low revs and load, with a prevailing lower ambient (barometric) pressure, boost pressure should be correspondingly lower than when ambient pressure is at nominal value.

The waste-gate mechanism I am familiar with is operated by bleeding off a sample of the boost from the compressor side of the turbo and pitting it against a calibrated spring/diaphragm combination, with a mechanical linkage to the waste-gate on the turbine side.

If for whatever reason - including lower ambient pressure, or a punctured compressor-to-intake manifold hose - enough boost is not developed at higher revs and load, the waste-gate stays put(shut).

Could it be that the "self-regulating" label assigned to the turbo-charged engine derives from this ?

 

[Lou Scannon]

The wastegate is not having any effect until it is open, at least a crack. Then it is acting like a simple mechanical pressure regulator, with more or less flatness vs flow rate according to the design. But it is referenced to atmosphere, so it cannot compensate for changes in atmospheric pressure.

 

[Paulista]

hemi,

The loaded-spring/diaphragm arrangement that comprises the wastegate should only budge if the pressure for which the former assembly has been adjusted/calibrated for is attained during operation.

So unless the latter value of boost pressure is attained, the wastegate will be shut and boost will build up.

To what degree the build-up of boost compensates for a less dense atmosphere, is a matter for speculation.

At 5 k feet, the pressure/temp is some 12,24 psia/ 41,7 ºF, compared to 14,69 psia / 59ºF at sea-level.

Max boost on my 4-cylinder turbo-diesels at full-load is some 1,3 bar (gauge) at sea-level.

Enough zoomp there to offset a 2,4 psia penalty at 5 k feet ?

 

[patprimmer]

To the best of my knowledge, a wastegate works as follows.

Both atmospheric pressure and spring force holds the valve closed and boost pressure acts on the other side to blow it open. Exhaust pressure difference will also act against the valve face, but the valve is normally a lot smaller than the diaphragm.

The boost pressure will be maintained at ambient atmospheric which is variable, plus the spring which is constant while the valve is on the seat, unless deliberately reset, so due to this, a turbo engine will normally drop of power at reduced atmospheric density, but it will not fall off by as much as an NA engine will, so it is partly but not fully compensating.

The magnitude of the partial self correcting will be proportional to the ratio of atmospheric pressure to spring pressure.

There will also be a small variation due to exhaust pressure on the valve, but it should be much smaller than the other forces involved, and will tend to increase with speed and with increased air density as that will produce more exhaust gas. That will tend to further diminish the self regulating tendency.

To further complicate it, the increased exhaust gas will require the waste gate to open further, thus increasing effective spring pressure and so offsetting to some degree the effect of exhaust gas pressure on the valve.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[Paulista]

patprimmer,
You have emphasised that it is not only the spring pressure, but also the presure of the ambient air that acts on the diaphragm that holds the valve closed.

The pressure exerted on the face of the valve seat in the exhaust assembly would seem to be a second-order effect.

However, unable to generate insight into why the amount of self-regulation is proportional to the ratio of ambient atmospheric pressure to spring pressure....Obviously, in the limit (a vacuum), zero self-regulation.

Now consider for sake of argument a standard turbocharged engine, less wastegate....would the self-regulation for lower ambient pressure be any better?

Wonder if someone could come up with a simplified block diagram showing the control inputs, outputs and control loops in a turbocharged engine.

 

[patprimmer]

I have yet to see a piston engine run in a vacuum.

It should be quit obvious that the wastegate is the regulating device, so removing it should remove any regulation.

If you have a spring adjusted to say hold 30# boost before the wastegate valve cracks of it's seat plus atmospheric pressure acting on the wastegate valve. If atmospheric pressure is say 15 psi, you get 45# boost. If it is 10 psi you get 40# boost

The atmospheric pressure dropped by 1/3 and if NA, ignoring effective total compression ratio, the power would drop by about 1/3.

The drop from 45# to 40# is only a drop of 1/9 or 1/3 the drop of an NA motor.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[stenis]

The reason for the dip in the torque curve may very well be a due to some kind of resonance in the intake or exhaust system. Variable intake or exhaust systems or adjustable cams may be the solution.

Some reasons behind dips in the torque curve and some possible cures may be found in the link below.

http://www.gtisoft.com/confarch/Torino.ZIP