Using Blair to spec a Cam

[statsystems]

Could someone explain STA in laymen terms so it is more understandable for general engine tech. I can make assumptions on the term but would probably be wrong.

Specific time area. Inverted radius comes to mind.

Thanks. Terminology is one of the things that creates a division in these discussions and most will not ask and I fail to ask when I should.

I have a 2 stroke book around here with a ton of STA math in it. Didn't think of it until now. I doubt it would be accurate for valved engines. But the principal is the same. And you have to account for the expansion chamber.

[hoffman900]

As Neels pointed out in the other thread, STA is proportional to the Discharge Ratio.

Think of it this way, a valve with a certain annular area and a certain discharge coefficient can flow so much mass. It’s a game of how much mass can you allow into a cylinder in a given amount of time.

The pumping phases are easier to wrap your head around. You need so much area, contingent on how well the head moves air/fuel, for a target BMEP, at a given rpm. So the pumping phases would be; on the intake from TDC to BDC. On the exhaust, BDC to TDC. So your duration is fixed. The higher your BMEP the more you need to trap, and the faster you spin it the less time you have. The others are not as easy as they would include the closing or the opening side of the valve lift curve.

[SchmidtMotorWorks]

There seems to be a lot of ways to play around with STA.

Could someone explain STA in laymen terms so it is more understandable for general engine tech. I can make assumptions on the term but would probably be wrong.

Here is what happens in a simplified explanation:

If you know:
The valve dimensions
The valve lift profile
You compute the flow area at the curtain at every given crank angle
If you do that for each degree of crank rotation you can write those areas down in rows in a paper-list, spreadsheet or computer program (computer program is easiest way to do it, much easier than people that have not tried think it is).

Then you separate the rows by what the piston is doing at that time in to regions such as,
Exhaust open to BDC (blow down)
BDC to TDC (Exhaust Pumping)
Intake Overlap
Exhaust overlap etc, see first page.

Blair divided it one way, you can divide it any way that you want to, the main point is that you divide all the activities into sets. Think of it as a dissected and accounted examination of your cam and valve combination, like a Dr and an accountant would do if they worked together. When numbers are listed and then added-up, that is called "integration" (as in put-together).

Once you have those sets of listed areas you add all the areas up (in each set separately) and you then have 6 numbers, one for each of the six ways the cam activities were divided. Those numbers tell you how much flow area you have for each of those activities. (You can probably see by now that you could divide that activities in other ways and accomplish a similar concept.) The point of the system is to make a way to make easy to compare measurements from engine to engine.

Knowing the flow area is a nice starting place but for it to be meaningful, you have to compare it to the performance level and time (by RPM) the area is open of the engine (Blair used BMEP, you could use anything you want). The higher performance the engine, the greater the flow,time,area is required, and the lower the performance the lesser flow,time,area is required.

All of the above is the concept for having some way to measure the cam and valve ability to feed an engine. This is just book keeping like an accountant does.

Now that we have a concept to measure with, anyone can use that concept (or variations of it like I did with CamFlowRPM using flowbench data instead of area) and make a list of the flow,time,area and BMEP for the type of engines you care about.

Once you have that list, you can make a graph that compares those values for all of the engines.
If those values graph out close to a straight line or a curve, then it is easy to write a formula that defines a line that approximately fits those points.

Do that for all of the 6 regions of cam activity and you will have 6 short formula like you see on the first page.
You can use that formula to set a target for an engine that you want to build in the future.
You will do the same computation for the engine you are planning to build, to see if it has a similar flow,time,area as other engines that are similar and you know run well.

If it is off, then you may need to change the cam of the valves to bring it in line.
The beauty of the system is it fully boxes in the cam and valve combination.

The hardest part for some people will be collecting a set of engines to use as a baseline.

For that, it is better to use one engine that is very similar to what you want to build, than a general purpose formula that has no way to know what you are trying to do.

That sounds like a lot of work, but if you make a computer program to do it, it is simple and easy after that.
Making computer programs isn't as hard as it used to be.

[GARY C]

There seems to be a lot of ways to play around with STA.

Could someone explain STA in laymen terms so it is more understandable for general engine tech. I can make assumptions on the term but would probably be wrong.

Here is what happens in a simplified explanation:

If you know:
The valve dimensions
The valve lift profile
You compute the flow area at the curtain at every given crank angle
If you do that for each degree of crank rotation you can write those areas down in rows in a paper-list, spreadsheet or computer program (computer program is easiest way to do it, much easier than people that have not tried think it is).

Then you separate the rows by what the piston is doing at that time in to regions such as,
Exhaust open to BDC (blow down)
BDC to TDC (Exhaust Pumping)
Intake Overlap
Exhaust overlap etc, see first page.

Blair divided it one way, you can divide it any way that you want to, the main point is that you divide all the activities into sets. Think of it as a dissected and accounted examination of your cam and valve combination, like a Dr and an accountant would do if they worked together. When numbers are listed and then added-up, that is called "integration" (as in put-together).

Once you have those sets of listed areas you add all the areas up (in each set separately) and you then have 6 numbers, one for each of the six ways the cam activities were divided. Those numbers tell you how much flow area you have for each of those activities. (You can probably see by now that you could divide that activities in other ways and accomplish a similar concept.) The point of the system is to make a way to make easy to compare measurements from engine to engine.

Knowing the flow area is a nice starting place but for it to be meaningful, you have to compare it to the performance level and time (by RPM) the area is open of the engine (Blair used BMEP, you could use anything you want). The higher performance the engine, the greater the flow,time,area is required, and the lower the performance the lesser flow,time,area is required.

All of the above is the concept for having some way to measure the cam and valve ability to feed an engine. This is just book keeping like an accountant does.

Now that we have a concept to measure with, anyone can use that concept (or variations of it like I did with CamFlowRPM using flowbench data instead of area) and make a list of the flow,time,area and BMEP for the type of engines you care about.

Once you have that list, you can make a graph that compares those values for all of the engines.
If those values graph out close to a straight line or a curve, then it is easy to write a formula that defines a line that approximately fits those points.

Do that for all of the 6 regions of cam activity and you will have 6 short formula like you see on the first page.
You can use that formula to set a target for an engine that you want to build in the future.
You will do the same computation for the engine you are planning to build, to see if it has a similar flow,time,area as other engines that are similar and you know run well.

If it is off, then you may need to change the cam of the valves to bring it in line.
The beauty of the system is it fully boxes in the cam and valve combination.

The hardest part for some people will be collecting a set of engines to use as a baseline.

For that, it is better to use one engine that is very similar to what you want to build, than a general purpose formula that has no way to know what you are trying to do.

That sounds like a lot of work, but if you make a computer program to do it, it is simple and easy after that.
Making computer programs isn't as hard as it used to be.

Dang your short answers are long. but thanks for the detail.

In simple (to begin) it is window area in relation to crank degrees?
And then mapping it through the rotation?

[SchmidtMotorWorks]

In simple (to begin) it is window area in relation to crank degrees?
And then mapping it through the rotation?

Yes, and then including target performance and engine speed.

Someone could make a phone app for it easy enough.

[GARY C]

In simple (to begin) it is window area in relation to crank degrees?
And then mapping it through the rotation?

Yes, and then including target performance and engine speed.

Someone could make a phone app for it easy enough.

When you say 6 regions of cam would that be the 4 open and close events, overlap and LSA?

[pcnsd]

In simple (to begin) it is window area in relation to crank degrees?
And then mapping it through the rotation?

Yes, and then including target performance and engine speed.

Someone could make a phone app for it easy enough.

When you say 6 regions of cam would that be the 4 open and close events, overlap and LSA?

From a previous post

Since Blair uses metric 177.645 = 12.2482 Bars

► BMEP_BARS = 12.2482
► (5.02 * BMEP_BARS + 57.78 ) / 10000 = 0.0119265964s/m (intake pumping)
► (1.7775 * BMEP_BARS + 74.822) / 10000 = 0.009659317549999999s/m (exhaust pumping)
► (4.1185 * BMEP_BARS - 17.985) / 10000 = 0.00324592117s/m (intake overlap)
► (3.0296 * BMEP_BARS - 11.363) / 10000 = 0.002574414672s/m (exhaust overlap)
► (1.6329 * BMEP_BARS - 7.1871) / 10000 = 0.001281298578s/m (exhaust blowndown)
► (2.4022 * BMEP_BARS - 14.57) / 10000 = 0.001485262604s/m (intake ramming)

[GARY C]

Yes, and then including target performance and engine speed.

Someone could make a phone app for it easy enough.

When you say 6 regions of cam would that be the 4 open and close events, overlap and LSA?

From a previous post

Since Blair uses metric 177.645 = 12.2482 Bars

► BMEP_BARS = 12.2482
► (5.02 * BMEP_BARS + 57.78 ) / 10000 = 0.0119265964s/m (intake pumping)
► (1.7775 * BMEP_BARS + 74.822) / 10000 = 0.009659317549999999s/m (exhaust pumping)
► (4.1185 * BMEP_BARS - 17.985) / 10000 = 0.00324592117s/m (intake overlap)
► (3.0296 * BMEP_BARS - 11.363) / 10000 = 0.002574414672s/m (exhaust overlap)
► (1.6329 * BMEP_BARS - 7.1871) / 10000 = 0.001281298578s/m (exhaust blowndown)
► (2.4022 * BMEP_BARS - 14.57) / 10000 = 0.001485262604s/m (intake ramming)

Thanks!

[SchmidtMotorWorks]

Thanks!

Yep, and the concept could be modified to measure overlap area in a different way or whatever you think might be more insightful.

It is a combination of two things:

1. A way to measure your combination.
2. A way to compare it to other combinations.

When it comes down to it, it is really quite basic.

[GARY C]

Thanks!

Yep, and the concept could be modified to measure overlap area in a different way or whatever you think might be more insightful.

It is a combination of two things:

1. A way to measure your combination.
2. A way to compare it to other combinations.

When it comes down to it, it is really quite basic.

It seems to boil down to a difference in terminology and methods.

[SchmidtMotorWorks]

Thanks!

Yep, and the concept could be modified to measure overlap area in a different way or whatever you think might be more insightful.

It is a combination of two things:

1. A way to measure your combination.
2. A way to compare it to other combinations.

When it comes down to it, it is really quite basic.

It seems to boil down to a difference in terminology and methods.

One of Blairs talents was distilling a problem down to the most compressed and complete definition.

But you have to consider that besides doing this empirical work, his most used and expanded on work is his 1D simulation work. People continue that because it is the most efficient way to get work done.
You can get a simulation of 5 manifold designs done at a good consulting firm for much less than you can print and test just 1.
These days, dyno testing is becoming just a confirmation step.

[vannik]

STA is not a new concept, the earliest reference to it I could pick up was in this book, but it might be older. It is just a lot easier to use with a digital computer today.

PHSchweitzer.JPG

So nothing new, just used in a better way.

[vannik]

STA is not a new concept, the earliest reference to it I could pick up was in this book, but it might be older. It is just a lot easier to use with a digital computer today.

PHSchweitzer.JPG

So nothing new, just used in a better way.

[Stan Weiss]

This what I have come up with. The number are off a little while the overlaps are out of the range I would like. Both lobes I am using are symmetrical. I think making them non symmetrical is what is needed. Not sure when or if I will have the time to do that as this is outside what the thread was started for.

Stan

Code: Select all

► (5.02 * BMEP_BARS + 57.78 ) / 10000 = 0.0119265964s/m (intake pumping)
► (1.7775 * BMEP_BARS + 74.822) / 10000 = 0.009659317549999999s/m (exhaust pumping)
► (4.1185 * BMEP_BARS - 17.985) / 10000 = 0.00324592117s/m (intake overlap)
► (3.0296 * BMEP_BARS - 11.363) / 10000 = 0.002574414672s/m (exhaust overlap)
► (1.6329 * BMEP_BARS - 7.1871) / 10000 = 0.001281298578s/m (exhaust blowndown)
► (2.4022 * BMEP_BARS - 14.57) / 10000 = 0.001485262604s/m (intake ramming)

===================

Intake BTDC (IVO to TDC) = 7.463
Intake Pumping (TDC to BDC) = 119.310
Intake Ramming (BDC to IVC) = 14.737
Intake Overlap (IVO to EVC) = 35.689

Exhaust Blow-Down (EVO to BDC) = 12.723
Exhaust Pumping (BDC to TDC) = 96.753
Exhaust ATDC (TDC to EVC) = 5.565
Exhaust Overlap (IVO to EVC) = 29.578

VALVE     Lift      Opens   Closes  Duration
                 Deg BBDC  Deg ATDC             Area
         0.00000    78.89 |  55.89 | 314.78 |  73.85
         0.00200    77.96 |  54.96 | 312.92 |  73.85
         0.00400    77.08 |  54.08 | 311.17 |  73.85
         0.00600    76.25 |  53.25 | 309.49 |  73.85
         0.01000    74.67 |  51.67 | 306.34 |  73.83
         0.02000    71.19 |  48.19 | 299.38 |  73.79
         0.04000    65.47 |  42.47 | 287.94 |  73.61
         0.05000    63.01 |  40.01 | 283.02 |  73.52
         0.10000    52.92 |  29.92 | 262.84 |  72.71
         0.15000    44.69 |  21.69 | 246.37 |  71.70
         0.20000    37.34 |  14.34 | 231.69 |  70.48
         0.25000    30.50 |   7.50 | 217.99 |  68.91
         0.30000    23.94 |   0.94 | 204.87 |  66.97
         0.35000    17.53 |  -5.47 | 192.07 |  65.01
         0.40000    11.21 | -11.79 | 179.43 |  62.77
         0.45000     4.91 | -18.09 | 166.81 |  59.79
         0.50000    -1.45 | -24.45 | 154.10 |  56.93
         0.55000    -7.93 | -30.93 | 141.14 |  53.79
         0.60000   -14.62 | -37.62 | 127.76 |  49.78
         0.65000   -21.63 | -44.63 | 113.75 |  45.41
         0.70000   -29.12 | -52.12 |  98.75 |  40.00
         0.75000   -37.40 | -60.40 |  82.20 |  34.18
         0.80000   -47.07 | -70.07 |  62.85 |  26.41
         0.85000   -60.14 | -83.14 |  36.73 |  15.64

[Vincenzo]

Stan, you are way ahead of me, I haven't finished with the exhaust side yet.

My first attempt to build a model of the Pontiac engine with STA values to match those of your original post, when simulted in EngMod4T revealed an extremely high pressure still existing inside the cylinder at time of intake valve opening, thus initial intake flow was seriously compromised. This situation was sucessfully addressed by an increase to the size of the exhaust valve, seat, throat and manifold, which resulted in STA values that were moved away from your originals. This with a symetrical valve lift design.

If this were all there was to solving this issue, things would be fairly straightforward, but not so.

Running a simulation of the complete valve train and components in 4StHead resulted in a dynamic at 7500 rpm which showed a very serious valve bounce condition at valve closing. This was overcome by a valve lift re design, where it was found necessary to change to a non symetrical design to have control of events and eliminate the bounce. This with a Titanium valve, the weight of a steel component was so detrimental that separation and bounce were astronomical.

Valve opening is always compromised in a pushrod engine by flex in the valve train, and in the case of an exhaust valve, the valve opening is accomplished only after the very high pressure within the cylinder is overcome, a result of which is flex and bending of the pushrod and rockers. This flex can result in the loss of duration to a considerable extent, and in the case of my simulation, 18 degrees delay in opening were computed.
Duration of the intake is also compromised, but the opening delay is very much less than that of the exhaust owing to the absence of the high in cylinder pressure, although some flexure of pushrod and rocker will still occur.

The STA values for a static situation might be very much easier to solve than the dynamic, but to make allowance for all the variation that occurs in a running engine might just prove to be a bridge too far !