Interesting read on the Power Split Device

We can never get too much of the PSD.

Tom

 

The Flash PSD demo is one of the best from that site.

 

Can this really happen ???

See the image. I think this diagram is wrong in the extreme values, which makes me doubt how correct it is in the "normal" values ...

Speed: 110 MPH
ICE: 0 RPM
MG1: -16,900 RPM
MG2: 6,500 RPM

Or some other wierd combination like going full speed then some in reverse:

Speed: -25 MPH
ICE: 4,500 RPM
MG1: 20,003 RPM
MG2: -1,463 RPM

 

'wheels of the bus go round and round, round and round'...

sorry, couldn't help myself with the Seinfeld reference to the "The Bet" episode where Jerry's trying to not, well, er...hopefully all the Seinfeld fan's will get it!

And on topic, the PSD is the heart and soul of HSD. Center of my universe, anyway.

 

Re: Can this really happen ???

They match pretty closely the results from the Prius MPG Simulator nomograph. Just because they're mathematically possible doesn't mean they're mechanically possible. For example, the car won't allow the ICE to run at zero RPM at speeds above 40 MPH (unless the driver drops it into neutral below 40 MPH on a downhill and gains speed). It spins the ICE during coasting or braking (with no fuel flow) at about 1000 RPM precisely to keep MG1 speeds from approaching such extremes. And the electric motors power the car in reverse, so you'll never see an ICE speed higher than idle speeds, if that.

 

It's not "mechanically" possible if you get flashing red lines AND if the rpm exceeds 10,000.

Otherwise, its all possible, for example the shifting into neutral + downhill acceleration trick to reach 50 mph with ice at 0.

 

Re: Can this really happen ???

Wrong how?

Looking at the values you chose for MG2 and ICE, I'd say that the MPH and MG1 look right to me. Essentially, if you shift into Neutral when the ICE is off and then coast down a long steep hill until you reach 110 MPH, you will get the -16,900 RPM indicated for MG1. This significantly exceeds the design specifications for MG1, so I'd advise against attempting it, but the numbers look right to me.

Additionally, you'd have to do some significant modification of the car to get the ICE to spin at 4500 RPM while backing up, but if you can find a way to do that while backing up at 25 MPH, then MG2 will be spinning at 20003 RPM. Again, I'd advise against making such modifications to your vehicle and attempting this, but the diagram is an accurate representation of the behavior of the Power Split Device.

 

Hello,
It seems you do not understand the nomograph. Only STRAIGHT lines will work. This is why the line turns red when you go "off the curve."
Think about it for a second - the PSD uses gears that are always engaged. There is no slippage. For the line you have posted up to REALLY occur, then you have just stripped the gears in the PSD! That is why the person who made that nomograph makes it flash RED when it's off the curve. You then know right away that combination of gearing is impossible.
The numbers in the display turn red also.

It is a simple nomograph.

Guy in WNY

 

I believe this isn't HSD it is THD.

 

Thanks for the clarification - I did not pay attention to the colors. The practical aplication is that makes grasping the full meaning of it not so simple , and I think your post can us some corrections as well.

The diagram appears correct and I may have been wrong in saying it was not (based on the "impossible" ranges it was depicting). And I could take back the "wrong" comment from my first e-mail now, but I'll leave it in it for reference as some of the subsequent threads will be meaningless without it

However, you are yourself wrong it appears. On two counts. The meaning of red lines, and if teeth will be stripped.

The line seems to begin turning red when stepping out of the *software limit* he refers to. Not when teeth begin to strip. Teeth will *not* strip at any point in the range of the diagram, red or green. Or at leat I think so (I have not bothered to calculate the precise ratios).

The "simple nomograph" comment applies here very well - the straightedge in the middle begins to curve because the value lines on the sides run out of range (and begin to blink red). He had to draw them much too long for the scale to work with a straightedge, so he draws them only to the "safe" limit and "bends" the straightedge and makes it red when out of the range shown on the sides. This does not mean *impossible*!

To help better understand how planetary gears work, let's look at a simpler example.

Try this, to see my point easier: keep MG2 at 0. Set ICE at 2,022. MG1 is now at 7,278. Nothing magic about the ICE #s, just some value that keeps the MG1 below 10,000 and above 6,500 while MG2 is at 0. The ICE (orange) can rotate as fast as you want and the only thing that would happen to the MG1 (blue) is that it would spin faster. No teeth will be pulled, nor there will be slippage indeed. The diagram seems to correctly represent this condition by allowing the ICE to MG1 RPM ration to remain constant.

Since MG1 can go up to 10,000 per the article, the scenario I give here is perfectly possible physically and without damage, most likely.

Unless I'm wrong somewhere, this is well within the theoretically possible and, as some have pointed out and the article itself states, not destructive combination (MG1 can go to 10,000 RPM per it, mechanically).

To put a practical prospective on this, the situation above would happen when you step hard on the brakes, the car does not move, but you press the gas pedal hard. You are effectively force-charging the battery in this case. I am not at all sure how the "software limit" would affect the operation of MG1 though - may be charging would stop or the engine would cut-off to keep the MG1 RPM down...

Anyway, Completely Different Topic Below.

This diagram makes a very nice demonstration about how the MG1 goes from + thru 0 (at about 65 mph per this diagram) to -. It is a very gradual motion if one is watching the speed of the car and that transition thru 0 should not be stressing the gear components IMO in normal operation. However, since MG2 is directly related to the wheels, I cannot but wonder what forces play in this PSD in the situation of a panic stop from say 80MPH where both front wheels lock. ABS will kick-in but not immediately I presume and therefore forces in the differential and the PSD must be substantial to cope with these abrupt changes in wheel rotation in the abscense of "soft" components like clutches or viscosity-linked traction components... Any thoughts on this anyone?

 

MG2's forces are not transferred through the PSD. MG2 is tied directly to the drive train, which in turn is tied to the ring gear of the PSD. Forces from the ring gear can be sent to the drive wheels, and forces from MG2 can be sent to the drive wheels. Regenerative braking forces go straight from the drive train to MG2 and are not transferred through the PSD. Inertial forces will exist in the PSD when the drive train changes speed, but these are minimal and should be compensated for by MG1.

Tom

 

I had no idea that such an elegant engineering marvel was under my hood!
Thanks guys. That's a BookMark.

 

Yes, that indeed force-charges the battery -- up to a point. As the battery approaches the 80% upper charging limit (enforced by the car's computers), RPM will drop even with the go-pedal to the floor.

 

The "drive train" from the PSD is a chain as far as I know, which does not have much "stretch". To me this means that both MGs are "hard-coupled" pretty much to each other and to the ICE and to the wheels... I'm curious what's saving (in the case of an instantaneous hard stop from say 80 mph) the MGs, the ICE, and the rest of the gears that are effectively hard-coupled to each other and to the wheels?

In a regular auto transmission, there is always slippage in such conditions, so the effectively 0 speed on the driveshafts when the brakes are locked does not immediately translate in 0 RPM in the flywheel and engine camshaft etc. With manual transmission, I think the clutch will also slip under the severest stress and in case the driver forgets to depress the clutch pedal during a panic stop (the engine dies in this case ultimately, but I think some slippage in the clutch plates would make it not instantaineous).

So, what's the equivalent of a "slippage" link in the Prius? I see nothing mechanical there. The only thing that would ease the severe shock from locking the front wheels in hard braking would be the ABS system IMO. Is there anything else I am not accounting for? I am talking when both front wheels lock suddenly at speed - if only one locks, the differential will transfer the rotation to the other wheel so not much would go to the rest of the drive train...

Thanks!

 

The MGs and ICE are always hard-coupled to each other, with a slip clutch located between the PSD and the ICE. The little clutch is used to minimize shock loading and vibration when the ICE starts, so it's not really relevant to your question. MG2 is directly part of the drive train. Think of it as a bulge in the drive shaft. It doesn't really care if it speeds up or slows down suddenly, as it is easily able to handle the forces. Since MG2, the ICE, and MG1 are always coupled through the PSD, when one changes speeds the other or others must also change speeds to compensate. The ICE does not change speeds quickly. This leaves MG1, but that is good since MG1 is a small, high speed motor-generator. If you cut the field current to MG1 it will freewheel with very little resistance or inertia. Even if you don't cut the field current, it's only an electric motor. Nothing gets stripped or broken if you force it to spin or stop it from spinning. It's not like stripping gears or throwing a rod. The forces in an electric motor are magnetic. That's where your soft coupling comes in.

Tom

 

Great link

Thanks

 

Thanks! I did not know there is a slip clutch and I thought that there must be something like that or things would break way too soon otherwise and there would be a lot more jolts felt during various conditions...

 

Hey, glad you guys like the flash demo I made!

I keep meaning to update it with the 10k RPM limit, and also allow manual sliding of MG1... I also want to add some buttons to set up the states that john1701a's site demonstrates just so you can see what's happening during different real-world driving conditions. As was pointed out above, I didn't write in any realistic mechanical constraints, just mathematical ones, so it's more of a "planetary gear set" demo than a "HSD demo."

Still, I like being able to drag the sliders beyond realistic limits just to see how crazy bad it would be for MG1 if you drove 80mph with ICE not spinning.

Side note, my PSD page jumped from like 50 hits/day to about 12k hits/day after it made the front page of stumbleupon.com for a couple days. Lots more people are interested in this than I ever guessed!

 

I learned a bit more about this recently. It's not really a slip clutch, but a torsional damper. Think of it as a springy coupling between the engine and PSD. It never slips, but it does allow for a bit of play.

Tom