Engine technology thread.

[Draco-REX]

Originally Posted by serialk11r

Are stock ECUs ever reprogrammable that way? I guess we'll have to wait and see what this car has. The more I think about it, the more I want Subaru to actually do what they suggested could happen for the STI, drop a few hundred pounds, raise the rev limit. This car modding business sounds like a total pain in the ass

Subaru's ECUs are very programmable. You rarely see a stand-alone in a Subaru unless there's some incredibly extensive modding being done. Hopefully the FT ECU will be the same.

Originally Posted by Grip Ronin

so are they planning to use toyotas variable valve timing? vvti i belive they call it

The FA20 uses dual AVCS (on both intake and exhaust cams). But that's essentially the same as VVTI. One profile with an oil pressure actuated cam gear to advance or retard the cam's timing.

[Dave-ROR]

Because of the D4S system I'd guess this is more Toyota ECM than Subaru..

[brufleth]

Originally Posted by serialk11r

Another question for people who might be able to answer it, what are the electronic limitations to raising revs? I imagine there may be issues with the ECU as they would only be programmed for stock rev range from the factory.

I actually develop "ECU" software (really FADEC software) for engines (albeit not car engines). The tables should be adjustable just like any other adjustment. So provided you can access the software tables and adjustments then you're good on the software side (either redistribute points on the table or just scale things differently).

The other concern is that the ECU and sensors support the higher speed and the resultant I/O. I would guess that you're probably fine because you're not going to jet engine speeds or anything.

Really it wouldn't surprise me if the tables and everything were setup for that kind of speed increase just because they would have over speed and endurance tested the engine beyond what will be fielded and probably just set the software and hardware up to handle that without significant modification.

[Draco-REX]

Originally Posted by Dave-ROR

Because of the D4S system I'd guess this is more Toyota ECM than Subaru..

I would expect that too, but I've heard differently. We'll see.

[serialk11r]

Thanks for the responses...
Can anyone answer the rod length question?

[Dimman]

Originally Posted by serialk11r

Thanks for the responses...
Can anyone answer the rod length question?

Is the situation that the ratio that gets slightly less friction also has less peak acceleration? Because then you can get a mass reduction with going to a lighter connecting rod. And then slightly lighter crank counter-weights, slightly lighter crank damper, etc...

[Want.FR-S]

Originally Posted by Dimman

Is the situation that the ratio that gets slightly less friction also has less peak acceleration? Because then you can get a mass reduction with going to a lighter connecting rod. And then slightly lighter crank counter-weights, slightly lighter crank damper, etc...

So just thinking physically from the engine, if the crank diameter does not change, the stroke does not change. Suppose if you do not change the blocks, by increasing the rod length, you need to change the piston head with less length or use a different piston head. Doing so, you also need to see if the longer rod length can clear the rotation to see if the rod will hit the wall surrounding the crank.

That is, if you do not change the displacement.

One thing I know, based on the calculation of mean piston speed, is that if you want to get high-rev, you need to reduce the stroke so that your rod will not break under high-rev. Motocycle and F-1 engine with high rev but pretty small stroke.

However, I do not know the benefit of going with a longer rod (or shorter for that matter). Can someone explain it?

[serialk11r]

Draw a diagram, you'll quickly see what's going on. Draw a circle with diameter equal to the stroke, then draw a circle with radius equal to the rod length.

An infinitely long rod would make the piston follow sinusoidal motion. A finite length rod will behave differently near the top and bottom of the stroke. The linear acceleration of the piston is higher in the top half of the stroke with a short rod, and slower in the bottom half of the stroke, while a longer rod has lower acceleration in the top half and greater in the lower half (roughly speaking).

Perhaps the more obvious difference though is the difference in the angle between the piston motion and the rod. At a greater angle, the force on the side of the piston is greater. Since the piston skirts and rings are responsible for half the total friction/fluid pumping loss in the engine, there is an appreciable decrease in friction with a longer rod, given the same masses everywhere. A longer rod probably weighs a little bit more though...

Here are some pictures that may help: http://en.wikipedia.org/wiki/Piston_motion_equations

My question is, what is the primary reason for going to a longer rod length?

[Dimman]

^ Just throwing stuff out there. What about combustion pressure and leverage differences?

[Want.FR-S]

Originally Posted by serialk11r

1. An infinitely long rod would make the piston follow sinusoidal motion. A finite length rod will behave differently near the top and bottom of the stroke. The linear acceleration of the piston is higher in the top half of the stroke with a short rod, and slower in the bottom half of the stroke, while a longer rod has lower acceleration in the top half and greater in the lower half (roughly speaking).

[Point 1] What are you talking about in bold sentence? The piston should be in linear motion. What does that have to do with infinite long rod? Even if you have the infinite rod, the piston should be just moving linearly back and forth because it is limited by the engine block and the o-rings.

However, if you chart the crank angle vs piston position, you see a sinusoidal graph. Is there any correlation with rod length? It seems to me that by increasing rod length the curve is moved upward on this diagram (re. chart shown in the wiki link). The amplitude is changed when you change the crank diameter (stroke).

Originally Posted by serialk11r

Perhaps the more obvious difference though is the difference in the angle between the piston motion and the rod. At a greater angle, the force on the side of the piston is greater. Since the piston skirts and rings are responsible for half the total friction/fluid pumping loss in the engine, there is an appreciable decrease in friction with a longer rod, given the same masses everywhere. A longer rod probably weighs a little bit more though...

How would you correlate the rotating angle between rod/piston with the force on the side of the piston? Given the same piston head, the surface area between the head and the block should be the same. The friction on that surface area should remain the same. With longer rod, the rotating angle at the piston side is decreased. So one can argue that the total *work* required to rotate the rod at the piston side is reduced with longer rod, but let's not confuse that with friction. The friction is based on the surface area and the content between the surfaces.

So without considering the increase weight of longer rod, with the same piston head and stroke, the engine with longer rod will have less rotating angle, i.e. less work required to rotate the rod. Thus the total work required to move the piston to rotate the engine is less compared with the one with shorter rod.

However, with increase weight of longer rod, the total *work* required to move the piston would increase due to the higher weight. The question becomes: would the effect of reduced rotating angle at the piston side, thus less work to rotate rod, *greater* than the penalty of increase weight of the rod?

On the other hand, based on the formula shown in the WIKI page, I did an quick Excel spreadsheet to show the difference with different rod lengths.

With a 6" rod and 2" radius (half stroke), the rotating angle = 38.94 degree, and the absolute maximum values of velocity shown is 2.11 at 75 degree (I was using 15 degree increment). The theoretical maximum velocity should happen at 73.17615 degree. The maximum angular acceleration happens at 0/360 degree with -2.67.

If I change the rod length to 7 inch with the same 2" radius, the rotating angle becomes 33.2031 degree, and the absolute maximum velocity is 2.08 at 75 degree with maximum angular acceleration at 2.57.

So with these values, one can argue that by increasing the rod length, the maximum angular velocity and acceleration is reduced so that less stress to the rod "if the mass of the rod does not change". However since longer rod = more weight with the same material, one should be careful not to make quick statement simply based on one or two factors.

[serialk11r]

I refuse to think about leverage differences, too complicated :P I prefer to just assume that metals have low hysteresis and that crank bearings have lower friction coefficient than piston skirts. Maybe someone can provide empirical evidence here.

I suspect combustion pressure is the biggest difference. A long rod gives more time between any given distance travelled around TDC, so your peak pressure is higher because more of the charge is burned by TDC, and the charge starts burning further up the compression stroke, so the piston doesn't push against a burning charge for as long. The net effect should be like an increased compression ratio in terms of cycle efficiency, without a higher physical compression.

Want.FRS, look at the Wikipedia article I linked you. In that case, a picture is probably worth a thousand words.
The piston's path is "linear", but its motion is quasi-simple-harmonic (aka sinusoidal) with respect to time. The rod adds some complexity to this, most obviously decreasing it around TDC. The Wikipedia article varies stroke and keeps rod length constant, if you change the rod length instead, you'll see that it's increasing BDC peak acceleration. The reason is the piston's postion at a given crank angle is the vertical distance between a circle of radius = stroke/2 and a tangent circle of radius =rod length, vertical meaning perpendicular to the common tangent of said circles. An infinitely long rod would make the piston's motion perfectly sinusoidal.

The rod transmits any force to the piston at the angle it makes with the piston's travel. The force on the side of the piston is completely determined by rod angle and force that the rod transmits. Increased force on a sliding surface increases frictional force.

[Want.FR-S]

Originally Posted by serialk11r

Want.FRS, look at the Wikipedia article I linked you. In that case, a picture is probably worth a thousand words.
The piston's path is "linear", but its motion is quasi-simple-harmonic (aka sinusoidal) with respect to time. The rod adds some complexity to this, more specifically increasing acceleration around TDC and decreasing it around BDC. An infinitely long rod would make the piston's motion perfectly sinusoidal.

Hmmm.. the graph shown on the wiki page: have you checked what the unit is on the x-axis? It is the angle of the crank, not time. If you see the animation of any ricorpical engine, you can see the piston is limited by that linear motion. And that has nothing to do with the length of the rod. What the heck are you trying to blew?

And, what is the relationship of increasing acceleration around TDC of a piston with its motion with respect to time? In fact, with longer rod and everything else the same, the acceleration around TDC is *decreased*, not increase as you said.

[Want.FR-S]

Originally Posted by serialk11r

The rod transmits any force to the piston at the angle it makes with the piston's travel. The force on the side of the piston is completely determined by rod angle and force that the rod transmits. Increased force on a sliding surface increases frictional force.

Are you confusing about where the force is applied? We all know that the explosion in the chamber *push* the piston, and then rod to make the crank rotate. Now you are telling me the rod transmits force to the piston to make it move? That is complete opposite to this.

And what does that have anything to do with this: Increased force on a sliding surface increases frictional force. When one applied force to an object on a sliding surface, if the object is not moving, the applied force is equal to the friction. When you increase force, provided that the object is not moving, the friction increase. Once you break loose the static friction, the object moves and then it is governed by its dynamic friction, which is less than static friction. The force applied needs to be greater than this dynamic friction to keep object moving or the object will slow down.

This is basic physics. What does this have to do with increasing rod length?

[serialk11r]

Originally Posted by Want.FR-S

Are you confusing about where the force is applied? We all know that the explosion in the chamber *push* the piston, and then rod to make the crank rotate. Now you are telling me the rod transmits force to the piston to make it move? That is complete opposite to this.

For one thing, Newton's third law, forces are mutual. Only during the expansion stroke does the piston see a net force downward, in every other stroke yes in fact the rod is what is moving the piston.

[second paragraph was irrelevant, not going to say anything about it]

This is basic physics. What does this have to do with increasing rod length?
The piston's net acceleration is in a straight line, but the rod transmits force in the direction of its length as the wrist pin and crank pins are free to move, so there is almost always a sideloading force. A longer rod decreases the angle at any crank position, and decreases the total force. This is basic physics/geometry.

Okay I will attempt to address this post. Responses in bold.
Oh sorry, just noticed, the first post does point out a valid mistake, sorry about that, now it's fixed.

Crank angle and time are interchangeable at constant speed, please don't get nitpicky now. And sorry I couldn't resist being smart on that last comment, but please...