Understanding mediocre power output with Dynosim

[mlody]

Hi!

A colleague of mine built a 2.5-liter (103 mm x 76 mm) type 4 engine with modified 1.8-liter heads, larger valves (intake around 46 or even 48 mm, exhaust around 42 mm - I'm waiting for more precise info), a compression ratio of about 10.5:1 and a wider throttle body of 50 mm. The camshaft has a duration of 239° at 1 mm lift, a total lift without rockers of 9 mm, LC of 106°, IO 13.5° BTDC, IC 45.5° ABDC, EO 45.5° BBDC, EC 13.5° ATDC. The engine runs with an OEM-style intake and exhaust from a Porsche 914 and a programmable efi system that also controls the ignition timing.

This combo produced a somewhat disappointing 201 Nm (148 ft-lbs) at 3400 rpm and 118 HP at 4700 rpm at the crank on a real dyno.

The engine also responds poorly to ignition timing - it starts knocking around 3500 rpm when the ignition goes above 23.5° BTDC, reaching about 25° at 6000 rpm.

I suspect the selected camshaft is too “soft” for the CR and valves. However, since I'm not an engine builder or tuner by any stretch of the imagination - just an enthusiastic hobbyist - I've decided to get Dynosim 6 and simulate the engine to better understand what's happening. Does anyone use this software and/or have experience with it?

I started with a stock 2.0 Type 4 and was able to satisfactorily reproduce the stock dyno chart with the Dynosim's default head flow values, but when I use custom flow values from Stan Weiss' flow table (e.g. Raby's LE200 heads), I'm having trouble matching even stock performance.

Perhaps someone has tried to simulate a Type 4 in Dynosim software and can share their experience of what to look out for? Or say where the problem of the engine's somewhat meagre power output could lie based on the above hardware data?


Regards,
mlody

[bugguy1967]

Were the heads ported? In order to make power, the cross sections of the entire intake tract need enlarging.
_________________
"A petrol engine can start readily, run smoothly and give every appearance of being in good order, without necessarily being in good tune." - Colin Campbell, "The Sportscar Engine"

[modok]

Some people will say you need to know what the heads flow, what the header can flow, ect.
I might disagree.
Possibly just reading the manual for how to work a flowbench will probably tell you most of what's wrong with it.
Most of what a flowbench is for..... is measuring the efficiency of flow, as in, how much it flows compared to how big the passageway is.
So to do anything you have to know how big the hole is.
And guess what, the flow is very much related to the size of the hole.

Like, just simple formulas or pipemax can tell you if the sizes of the valves, lift, ports, runner, header.... are in the ballpark for what you want and eachother.

Stock induction and exhaust system with HUGE valves...... raises a big red flag for me right away, just common sense but also I know, or, I have been told by several experts, that porting stock t4 heads large enough to match those valves is very difficult or perhaps impossible by itself.


A larger valve will only lose overall performance if the restriction is elsewhere.
Why? well, that's a little more hard to explain and that's where a sim might help, or might just confuse the issue, or also just looking at a lot of real world data should see the same trend.

If the valves aren't the restriction the shape of the power curve ends up reflecting whatever flow capabilities the induction system and exhaust have, and very prominent are whatever harmonic swings are happening in the intake and exhaust systems, because the intake and exhaust are then free to do whatever it is they naturally do for better or worse, and what they like to do is resonate differently at different speeds.

On the intake side sometimes a larger valve can flow less if the flow "stalls", which almost always is the flowpath expanding too fast and flow detaches from one side. While it's pretty easy to accelerate air it's a lot harder to slow it down again in an orderly fashon, like, why a venturi has a long tail.

If the valve curtain area is the major restriction, then the power curve ends up a smooth arch, as valves are actually in control of the flow. This is less exciting, but at least predictable.
In the middle is the happy zone.

I guess what I'm saying is the big valves aren't restricting the peak power, but they may be responsible for the engine tending to knock at certain RPMS, if the actual VE curve looks like the rocky mountains, the computer will be confused, everything will be confused.

[Alstrup]

Dont know the Dynosim. but 10,5 CR with a cam of that characte, with that bore is WAY too much. at 23-25 degrees timing you will get something like a 60% burn rate before 20 degrees atdc. you need less static compression or twin spark to pull that off.
_________________
https://www.thesamba.com/vw/forum/viewtopic.php?t=435993

[vwracerdave]

What octane fuel are you running? My gut feeling tells me you're not running enough. With that cam and 10.5 CR you probably need more that pump gas.
_________________
2017 Street Comp Champion - Thunder Valley Raceway Park - Noble, OK
2010 Sportsman ET Champion - Mid-America Dragway - Arkansas City, KS
1997 Sportsman ET Champion - Thunder Valley Raceway Park - Noble ,OK

[mlody]

Thank all of you for the responses!

The heads are supposed to have been ported, but I assume it was more of a smoothing than enlarging job. I'm still waiting for the exact info on what was done. From what I've found, the flow of the stock type 4 heads should be good enough for about 130 hp, therefore the smoothing job assumption above... I know for sure that larger valves are installed.

The idea behind the original intake and exhaust is legislative on the one hand - the car is registered as a classic, and the owner wanted to keep the look. However, I know there is a second engine with similar specs but a stock crankshaft. It was built first, and after the initial disappointing dyno runs, the owner assumed that the intake and exhaust were limiting performance. So the oem intake was modified with larger runners and larger diameter exhaust pipes were fitted, but this didn't make much difference - the engine made about the same power on the dyno. Thus, the second engine (the one I'm "playing" with) is running with stock parts.

I have the VE and ignition tables from the stand-alone ecu. According to the dyno plot, there's a healthy torque hump from 2800 to 3800 rpm and that's where the VE table goes over 80% in the ecu.

The engine is running with euro 98 octane fuel, so it should be equivalent to "super premium" in the US.

In general, the engine pulls nicely, runs smoothly and is quiet economical (does about 25 mpg), but above 5200 rpm it runs out of puff and there doesn't seem to be much you can do to change this as it starts knocking with more ignition... That's the reason why I'm trying to simulate the behaviour to check if changing the cam would be enough or if new heads with less CR are needed too...

Regards,
m.

[Alstrup]

OK.
Opinions vary. But.
First. That engine is totally misconfigured for what you- or he is trying to do.
Next. Even a stockPorsche 914 2 liter (4 cyl) runs out of steam at about 5200 rpm. So thats no surprise.
A stock type 4 muffler of today is 90% certain a JP Group version. The stock mufflers SUCK! Period! However, they can be made to work well if you are ready to cut them open and change a few things inside, but don´t expect a lot since the next party killers are the heat exchangers. Stock heat exchangers will not support more than about 120 hp.
If you are using stock intake parts from a 2 liter bus engine you will choke the engine pretty soon. anything more than about 100 hp and the system is totally stretched out. 1,8 intakes are the best and can support up to about 130 hp within reason. The bottle neck is the plenum and the throttle body. Anything more than 130-135 hp tops and you need custom intake runners.
But stop trying to make that turd run well. Get the obvious issues fixed first. THEN you can begin tuning.
Good luck.
_________________
https://www.thesamba.com/vw/forum/viewtopic.php?t=435993

[mlody]

Hi,

it's a 914, so he uses the parts of a 2.0 GB engine. But these are still stock parts, so they limit the flows...

I did a little more digging - the camshaft seems to be a Schleicher 324/106. From what I've found on the net, it should actually deliver quite decent performance values...
So I inserted it into the virtual world of Dynosim and then compared it with a Web 86a without changing the other parameters of the simulation. No matter what I do in the simulation, the Web 86a always performs significantly better from ca. 3000 rpm upwards. Then I tried the sim with a Web 86b, the change to the Web 86a was as expected (shift to higher rpm), but my sim didn't respond well - the values I had set for the intake and exhaust flow were too low to get an 86b to work...

Anyway... Have any of you compared a Schleicher 324 cam with a Web 86a (or 86b) on a real dyno and with the same engine and can share your experiences? Is a Web 86a really "more powerful" in the upper rev range than a Schleicher 324?

Regards,
m.

[Alstrup]

No. The 86a is not an rpm cam. In fact i don´t know what to call it. Not one of my favourites.
The 324 - is - an rpm cam, but the lift is relatively low and the lobe separation tight. The program can probably not compensate for that.
On the other hand, what do you get for $40.
If you want a program you can rely on to the extent of the accuracy of the data inserted. you need to spend significantly more money, like around 140 for a very basic program, and $500 for a good program and 5000 for a professional one.
_________________
https://www.thesamba.com/vw/forum/viewtopic.php?t=435993

[mlody]

What would be your preferred cam for a powerful but street-able 2.5-litre engine?

I absolutely agree with you that the results depend heavily on the input data and that you get what you pay for. But one has to start somewhere. Getting started with professional software might be a bit difficult for a beginner, apart from the cost. I'm just a keen enthusiast who wants to learn, and a sim appeared like a cheaper way to learn than building real engines and testing them on the dyno... I would love to build my own engine one day, but unfortunately the circumstances are not there yet.

Regards,
m.

[Alstrup]

You are going at it the wrong way. (but you are not alone)
ESPECIALLY with a type 4, and ESPECIALLY with a big bore, you need to start with deciding which rpm band you want the power in as well as the power level, and subsequently which exhaust you need to use to support that. Then you chose heads, THEN you chose the camshaft.
There are of course more than one road to Rome, but you want to chose the less bumpy one so to speak.
Also, with big bore engines. They were never my cuppa tea, i prefer smaller more efficient bores. But, it has recently become painstakingly clear to me that big bore type 4´s DO NOT like main stream camshafts if you want to pull decent power out of them. You want to look in the direction of RAT or Nowak cams. I will not make recommendations to which as I am probably not the right person to do that. Anything less than 98 mm bore, thats my game
_________________
https://www.thesamba.com/vw/forum/viewtopic.php?t=435993

[modok]

It's easier to learn from success than failure.
Study engines that work well, and then compare with ones that work poorly.

[mlody]

In my search for an understanding of the meagre power output, I stumbled across another mystery about the engine...

The colleague wanted to do an oil change at the weekend and check the valve clearance. With the valve covers removed, to further minimize the unknowns, we decided to measure the camshaft. According to the data we had received from the engine builder, the cam should have 9 mm lift. We took a dial indicator and measured on the pushrod side of the rocker arm. To our surprise, we only measured 6 mm lift. We tried both inlet and outlet valves on different cylinders, but every time measured 6 mm.

Is that possible or have we made an error? As I understand it, the lift on the pushrod side of the rocker arm should be about the same as the cam lift, but maybe I'm missing something...

Additional information gathered since the last post: The cylinder heads are fitted with 48mm intake and 38mm exhaust valves, are ported and have been apparently even tested/measured on a flow bench. However, there is no measurement data (yet). The combustion chamber has approx. 60 cc and the deck height is 1.8 mm, resulting in a compression ratio of 9.4:1.

Regards,
m.

[modok]

Oh, that's a clue.
The rocker ratio is about 1.3, so 6 at the cam should be 9 at the valve

[Alstrup]

[nogoodwithusernames]

I don't have any real life experience to add, but peruse some of Raby's old forum for lots of good info.
https://shoptalkforums.com/viewtopic.php?t=109178
_________________
71 Squareback, 1.7l T4 w/ crank trigger fuel and spark
(Conversion thread https://www.thesamba.com/vw/forum/viewtopic.php?t=732508&highlight= or https://shoptalkforums.com/viewtopic.php?f=1&t=151375&sid=f0542d44a322d290c29d6609fac7f215 )
74 914, 2.0L w/ D-Jet

-Derek

[mlody]

[Zed999]

[Alstrup]

I think we have touched on this before, but from a burn perspective, 3,5 mm is actually better than 1,8. 3,5 mm and the chamber acts like a semi hemi chamber and requires quite a bit of timing. 1,8 just leaves too much unburned, or burned too late, but still wants 34-36 degrees timing do do an even decent burn.
_________________
https://www.thesamba.com/vw/forum/viewtopic.php?t=435993

[SGKent]

If you have knocking from the higher compression, there are generally only two solutions in an older engine without all the knock sensors etc., they are use a higher octane or reduce timing. One disadvantage to the type 4 engine is that the timing scale is bolt on and can be off by many degrees, unless someone figures EXACTLY where TDC is and sets the scale to that. My first advice would be to use 110 octane racing fuel in that engine and see whether the knocking issue goes away. Hot spots on a head or piston can also trigger it. Running it as a diesel like it is will only destroy it. Rings are sensitive to being pounded on.

The type four engine typically makes max hp near the redline, which is 5400 rpm with a stock 914 European GA cam. That GA engine put out 100 HP as a 2L engine. So at 118 hp your friend isn't getting anything special with their 2.5L engine.
_________________
"Most people don't know what they're doing, and a lot of them are really good at it." - George Carlin