ESE's works engine tuner

[seattle smitty]

To Frits, who asked about it, outboarders do try to get the driveshaft between the engine and the lower unit to absorb some of the hit (without itself breaking). Some guys are trying plastics. A problem here is that to keep the C/G as low as possible you want to carry the powerhead as low as possible. The device holding the powerhead and the lower unit, and affixing both to the back of the boat, is called the towerhousing, and it has gotten ever-shorter over the years . . . with the result that the driveshaft (running inside the towerhousing) has gotten quite short . . . good for transmitting power but not good for absorbing shock-loads. This is why I'm wondering about 90degree twins and fours, and hoping to hear more about their pros and cons (relative to my application). I know that if you "had your druthers" you'd rather have exhaust systems that are entirely in a straight line out of the port; you get this with opposed-twins but you'd have to have your header pipe take a 45degree turn with a V-twin. But other than that, what do you see? I see a lot less abuse to the lower unit (and prop, and the "shear-pin" for the prop).

Thanks to all who are advising me here!!!

[seattle smitty]

Frits, I just took a first look at your link. Dang, it seems that the more I learn, the less I feel that I really know!! How interesting that at around thirteen-grand, the con-rod/piston/bearings suddenly are loaded in tension ATDC. And here I always had imagined that in a 2-stroke this tensile loading only happens when you suddenly let off the throttle. Have to go back to the link and see if I can pound this new idea into my old brain. Thank you very much, I think

(In thinking about your description of the actual power stroke, where the piston really is putting power to the crank, my first reaction is that here (this part of the stroke in a very-high-rpm 2-stroke) is where a long rod would really pay off . . . )

[TZ350]

Page 1040 is about the basics of pipe design and how to influence where the point of maximum depression occurs.

There are other collections of pipe talk on pages 620 630 640 650 660 670 680 690 700 730 740 which have un edited collections of raw material. If your interested in expansion chambers they are worth a look.

The pipe wall temp makes a huge difference , and much more accurately represents the real world wall temp average over the pipes length.



When you have temp varying with rpm, this overrides the temp at max power entry, and I found in Beta testing that 325 and 425 worked very accurately.

In a situation like the Aprilia model where a powerjet is switched off past peak, 450 is needed at peak rpm to get correlation but is too hot for our "normal" sims.

Well made pipes that go around a corner don’t loose any power to a dead straight one - but dead straight pipes do actually cause deto, and need a deto button of some sort in the header to suppress this artefact. I posted a pic of one of these on here ages ago.

When unleaded became mandatory in 98 most teams had a hell of a time keeping deto under control, having to run very rich and retarded.

The cylinders with the straighter pipes always gave a lot more grief, until a lowly dyno operator at Yamaha discovered that when he stuck a large pressure probe into the header, the deto went away.

They managed to keep this secret for a couple of years but when teams like WCM got hold of the bikes, and left pipes lying about as you see in the pics,word soon got around.



The "things " intruding into the header take varying forms,but all do the same thing - the usually bottom "bent "pipes dont have the same issue, so dont need the so called deto buttons.

I have a couple of deto buttons here but cant be bothered looking hard.

All they are in the form used by Yamaha and Roberts etc is a sort of mushroom. Or a penny on a stick. The penny being around 20mm dia by 3mm thick sitting on a stalk big enough for an M6 down the middle and a couple of flats to hang onto. The length being so the penny sits around 1/2 way into the pipe dia.

This thing situated about 50mm from the slip joint fixes the deto issue with a straight pipe.
Any bend that isolates the main body of the diffuser from the port seems to work.

The only plausible explanation I have is they shield the duct from infra red heat radiating from the pipe body interior - that overheats the plugging vol sitting in the duct. Goes along with not using heat wrapping on a header as it instantly causes deto. And as we well know Mr Thiel is a big exponent of keeping the duct charge volume small and cool - works for me.

I did proper back to back testing of the bent Vs dead straight pipes when building the BSL500.

The top pipes could be dead straight with just a small angle where the slip joint fitted onto the header,and these 2 pipes would deto as soon as you got close to spot on tuning.

Fit a deto button and the effect simply disappeared - we always had egt probes fitted and they made no difference.
On the bottom pipe of the tripple, this had a serious S bend to get the belly forward enough for it to fit nicely into the banana arm.

I found that as long as we had about 100mm of dead straight header fitting onto the slip joint I could get identical Hp within 1/2 in 55 as long as compensation was made for weld shrink on every joint, as well as every joint being hammered smooth. And the S bent pipe did not need a deto button at all.

I discussed the straight header effect with the Yamaha R&D chief at the time called "Suzan" and he opined that if you had a bend too close to the slip joint especially if it was bent in any way than on the vertical plane, this gave rise to an asymmetric plugging pulse flow, and this adversely affected the loop scavenge regime. Sounds pretty plausible to me.

- you need to get the latest version and always input the values of 325*C at the bottom of the usable powerband and 425*C at peak Hp, say 9500 and 13500 in the RG project.

The RG50 sim I am doing there is still good power past 14000 with a 735 pipe length. And this setup is to try to get as wide a power band as possible for kart track use with no PV or electronic PJ available ( yet ).

Baselines to remember here are that an Aprilia with 200* and 800 pipe revs to 14500, but needed an electronic powerjet to do this.

A 125 ICC kart engine with 84* and a 760 pipe will rev to 14500 + even with a flat line ignition, 30mm carb and no electronic powerjet. But both of those scenarios are very well worked optimisations of the parts available.

To get a deeper depression later in the cycle this is why we have a 3 section diffuser with a relatively shallow front end, then a long steep main diffuser leading up to the mid.

A short section of shallower angle leading into the belly helps to get a steeper angle on the main diffuser - closer to the mid at around 66%.

So - as a general rule the closer we have the steep section to the mid - the later ,and lower ,in the cycle toward TPC the depression will act.

But if we have the diffuser lengths correct the deep part of the depression will always move from closer to TPO under the peak, then ideally is centred on BDC at peak power rpm, then moves toward TPC
in the over rev area.

Thus - if you feel you need to move the max point closer to TPO you need to move the steep section closer to the header.

In regard to the flow in a 2T transfer duct there are several points you must realise about what is actually happening.

Firstly no matter what the porting arrangement (within the bounds of what we call normal ) there is always more pressure in the cylinder than in the case at TPO.

Thus there is backflow from the excess blowdown pressure at initial opening. This is what makes port stagger work - the port open first flows last.

As the piston continues to drop, the pressure ratio finally goes negative as the diffuser starts to do its work as the piston approaches BDC.

OK Wobbly, I have to ask this, even if the answer makes me cringe:
Why wouldn't I use a Vevey system in my 50 and 91cc buckets that need lots of midrange rather than peak HP?

Because the system reduces the peak power dramatically, and the only applications where the huge increase in low rpm power and bandwidth is of any use whatsoever is direct drive - no gears.

The KT100 needs to be able to haul off a hairpin at 8000 rpm and then spin to 16,000 to be able to only use 1 gear. In this case the torque multiplication of the very short gearing enabled by the peak rev capability makes the system workable.

With a bucket ( basket ) case you have a pile of gears to play with and only need to find the best compromise between powerband width and peak, to enable the least amount of gearchanges per lap.

Killing overev is easy, wind in some advance past peak,crank up the com, use a very steep rear cone, use a tight stinger,use a combination of port timing and pipe length
that has NO superposition resonance past peak.

All will make the powerband drop heavily after peak power, either by lowering the pipe temp or reducing pipe/port efficiency.

The problem with a long slow angle rear cone is that it returns a pulse ( though of low amplitude ) over a very wide band.

Thus in effect it "affects " the engine even more when well outside the tuning range of the mean reflection point than a shorter steeper design.

Taking this concept to the extreme, a flat plate "works" over a super narrow range ( and even then its not that effective due to the super short duration period ) and does virtually nothing outside the TL rpm.

A shallow diffuser does not create much negative pressure ratio around BDC, thus does not over scavenge the poorly configured transfers.

Once you have a well functioning sim of your engine and have run the Turbulent Entrainment model for your combustion Wiebe parameters the data feedback from TubMax is super accurate.

Getting the advance and stinger dimensions on the money to have this figure sitting on mid 900*C without WARNING - DETO - WILL ROBINSON screaming at you from the screen.

There are some very old guidelines that still work remarkably well today. Take as an example the RG50 T port cylinder I am working on at present. It has a 75% duct exit area at the flange = 24mm oval area. Take 0.58 of this diameter and we have 14mm, what is in the sim right now - 13.7mm with 18Hp at the crank. Thankyou Kevin Cameron.

A pair of transfers is a nightmare. I did an old 390 aircooled Husky and the only way to get any power at all was to re-liner and divide the HUGE transfer pair in two.

And for sure you cannot even think about using Frits or my design styles for what is essentially a lawnmower due to the pipe over scavenging the transfer system.

This means that at BDC a good diffuser (read fat) will create so much negative pressure ratio, most of the uncontrolled transfer streams will do a U turn and exit the cylinder.

When it comes to the piston port, the only approach I can see is to create a port as wide as you want, then use as little timing as is needed to match the STA to the blowdown and transfer numbers.

These small size (50cc) pistons can easily go 80% of bore with the Exhaust, so the limiting issue is going to be how much transfer you can create.

My first guess would be 60* back and 30* up, that at least will create some sort of loop effect for you.
Is it possible to get a reed into the inlet, then a boost port can be added and this works well, going all the way back to Kaden et al.

Page 950 .....
Its all about Exhaust blowdown...........
Pipes need to suck deep at BDC.....
Attachment 297806

Attachment 297808

Pay to follow Frits links to read the whole story.

Attachment 297809

From memory the plug tip also needs to be about 6.5mm from the piston crown.

[wobbly]

TeeZee,the older version used a completely different pipe modelling code.
In this setup we found that 50* gave the best correlation to dyno tested powerbands, and was a source of real frustration for Neels
who hates " fudging " numbers with a passion.
He came across a new thesis that modelled an RSW and this included a new pipe simulation technique, and this has been written into the new code.
Now the pipe wall temp makes a huge difference ,and much more accurately represents the real world wall temp average over the pipes length.
When you have temp varying with rpm, this overides the temp at max power entry,and I found in Beta testing that 325 and 425 worked very accurately.
In a situation like the Aprilia model where a powerjet is switched off past peak, 450 is needed at peak rpm to get correlation
but is too hot for our "normal" sims.

Re using the driveshaft to flex and absorb some of the rotational shock load - in a World Champs winning 1080cc twin SeaDoo making over 230Hp
I designed a shaft made from a special steel grade ( no I cant remember ) that was only 12mm dia.
This replaced a 35mm tube that ran thru behind the water intake grate and made a huge difference to the performance, and no it didnt break even with many
jumps out of the water where it hit the limiter then snapped the rpm down again on contact after a wave.

Well made pipes that go around a corner dont loose any power to a dead straight one - but dead straight pipes do actually cause deto, and need a deto button of some
sort in the header to suppress this artefact.
I posted a pic of one of these on here ages ago.

[dinamik2t]

Some 500 pages ago, dinamik2t posted the picture below.
Have any of you EngMod users managed to visualize the red/blue/green transfer port lines or is this just dinamik2t improving things in paint?
I think its a great feature if it was a way to add this lines within EngMod.

It's been a long time; I think I was trying to derive the angles from the blueprints of the APC cylinder Frits has provided back then.
This could be a more accurate representation of an APF (?) cylinder, having a straight horizontal front wall angle for the B tfr and it's rear wall flow direction towards the 1/6th of the bore on the vertical axis.
Total eff area is ~1302mm2, like stated by Frits on a pic of TFR/Blowdown angle-I think-areas.



Coloring is through paint.net and I agree that it would be a visually usefull add-on, if Neels ever has the time to implement it! I borrowed the idea from Frits.

-----------

I have a question: what sampling rate should be used when monitoring a knock sensor and an EGT sensor in a control unit making ignition timing corrections?

Furthermore, Wob could you give some directions on how to extend the depression of a pipe towards the rear side (towards TPO point)?

[husaberg]

Re using the driveshaft to flex and absorb some of the rotational shock load - in a World Champs winning 1080cc twin SeaDoo making over 230Hp
I designed a shaft made from a special steel grade ( no I cant remember ) that was only 12mm dia.
This replaced a 35mm tube that ran thru behind the water intake grate and made a huge difference to the performance, and no it didnt break even with many
jumps out of the water where it hit the limiter then snapped the rpm down again on contact after a wave.

.

As an aside Velocette used to design their bevel drives waisted for a similar reason.......from memory it worked like a torsion bar.
pretty sure I posted the Roarer story on the oldies thread. later maybe not.......

[AndreasL]

Like Wobs says, a thin shaft that flex.
I would have a look at "Orvar Supreme" from Uddeholm. Have heard about several successful applications when used in performance drive shafts.
http://www.uddeholm.com/files/PB_orv...me_english.pdf

Regarding the colored lines in EngMod, rumors say that they are in the making.

[Flettner]

torsional vibration, you want to be sure you are not designing a thin shaft ( spring ) that will hit a natural frequency and bust everything.

[AndreasL]

torsional vibration, you want to be sure you are not designing a thin shaft ( spring ) that will hit a natural frequency and bust everything.

Thats true, but that it takes some calculations to get it right was so obvious to me I did not mentioned it. Sorry if misleading anyone to belive it was a simple task.

[wobbly]

Re sampling frequency for a det feedback system - 12,000 rpm is 200Hz so to get instant response to a det event from a single cycle
this would need to be the sampling rate.
But this would probably be to responsive to random noise or det events, so 100 would be ample.

To get a deeper depression later in the cycle this is why we have a 3 section diffuser with a relatively shallow front end, then
a long steep main diffuser leading up to the mid.
A short section of shallower angle leading into the belly helps to get a steeper angle on the main diffuser - closer to the mid at around 66%.
So - as a general rule the closer we have the steep section to the mid - the later ,and lower ,in the cycle toward TPC the depression will act.
But if we have the diffuser lengths correct the deep part of the depression will always move from closer to TPO under the peak, then ideally is centred on BDC at peak power rpm, then moves toward TPC
in the overev area.
Thus to answer the question - if you feel you need to move the max point closer to TPO you need to move the steep section closer to the header.

[dinamik2t]

I was trying to generalize some power improvement-over-dimension changes on the 'default' HighPerf pipes and noticed that I was unable to create a deep trapezoid depression at peak, like you have described in the past as beneficial. Default dimensions (of either HiPerf or 3stage -and of the simple FOS) seem to always create a depression biased towards TPC. I will try these the next time, thank you for the explanation Wob!


ps. I 've been reading this http://books.google.gr/books?id=wC2d...page&q&f=false . The chapter (8) on Ignition is very illustrative if anyone is interested, describing among others the ignition process itself. I found interesting a comment on p.88, column2 : -talking about after-ignition events- "... The more mixture is consumed, the more heat is released, and this happens at the cube of the distance the flame front has traveled from the ignition source. In other words, the temperature rises more quickly as the flame front moves farther and farther from the point of origin. If this were to continue for very long, the remaining mixture would ignite simultaneously, causing severe detonation."

[TZ350]

Attachment 305313

This is going to be my interpretation of Seattle Smittys suggestion for header pipe finning.

More ideas for a exhaust header cooler.



A piece of finned pipe.



Or maybe one of these Bates motorcycle exhaust pipe clamps would be the way to go for header pipe cooling.

If anyone has an old Bates clamp to sell I would love to cut a deal.

[husaberg]

More ideas for a exhaust header cooler.



A piece of finned pipe.

Or maybe one of these Bates motorcycle exhaust pipe clamps would be the way to go for header pipe cooling.

If anyone has an old Bates clamp to sell I would love to cut a deal.

Those sort of things were very common factory on triumphs bsa's etc...

But I think DIY on a grandscale might be better and more potentially effective.

[TZ350]

I've observed that people here aren't bothered when two or three tech topics are being discussed concurrently, but I'm a little worried about interjecting.

Wouldn't worry, its a bit like a party, with different conversations all going on in the same room.... ... ... ...

[seattle smitty]

[QUOTE=wobbly;1130792247
Well made pipes that go around a corner dont lose any power to a dead straight one QUOTE]

Sigh . . . Well, that's what I had always understood for the last 45 years, but reading the Aprilia masters on the French site, I ran into one of them saying that a straight shot is best, or so I thought. If I can ever find it again, maybe I'll find he was only referring to the port itself. But then why would it matter for the port to be straight if it's okay for the header pipe and the rest to curve?

Talking about pipes, I guess everybody here is aware of Frits' new engine design, with ports as far as the eye can see. Has anyone asked how he's going to package it, given the two exhaust ports per cylinder? No problem on a single, but where do you put all the pipes as you add cylinders? Or do you run the headers from the two ports around into a single pipe? An American outboard tuner, Harry "Zak" Pasturzak, used to do this with Konig opposed-fours, connecting the top pair of cylinders (which amounted to a simultaneous-firing opposed-twin) to one pipe and the bottom pair of cylinders to another pipe (everybody else was connecting upper and lower pairs of cylinders (firing 180degrees apart) to a single pipe). (I haven't found a photo of Zak's pipes).