Manifold Vacuum

[Chris Duncan]

Flow dynamics question.

Using a TWM EFI 8 Tbody induction. Getting conflicting information about vacuum supply for the MAP sensor, FPRV (fuel pressure relief valve), and IAC (idle air control) valve.

The local Accel DFI dyno guy says ditch the existing vacuum junction and build a 50 cu in vacuum tank below the intake in the lifter valley that is supplied by 8 ct. 3/8" lines coming from each intake runner, and run the MAP, FPR and IAC all from that tank.

The guys at Kinsler say to keep the existing vacuum junction (2 cu in tank)(8 ct. 3/32" hoses) for the FPR and MAP and build another similar junction for the IAC valve, with separate 3/32" supply lines from lower down in the runners, in order to keep the 3 functions separate. They also say the 8 ct. 3/8" supply lines will conflict with idle characteristics.

The MAP and FPRV are dead end connections for vacuum signal only. The IAC essential adds air, in other words a vacuum leak. Obviously it will conflict with the MAP, FPRV accuracy if it's dumping air nearby.

The questions are would the large tank stabilize the vacuum enough to keep the MAP and FPRV accurate when the IAC is dumping air, and will the 8 ct. 3/8" lines affect idle?

For calculations sake I'm guessing the IAC has about a 3/8" opening max.

 

There is a good chance that I'm wrong, but here is my guess.

You want your computer to measure the pressure that your engine actually sees, IAC and all, so your MAP port should be the closest to the cylinder head.

Your fuel pressure regulator vacuum reference really does not matter. It just references the fuel pressure against the manifold vacuum. It should work fine without the connection, but it might go out of tune with elevation changes.

Your IAC is all that is left. Since you want your MAF sensor to have the most accurate reading possible, you should hook up the IAC as close to the throttle blades as possible.

So I think you want two vacuum junctions. They probably don't have to be huge. Use one for the MAP and one for IAC and FPR.

Lastly, you should look at a vacuum diagram for a fuel injected 5.0. It will have the relative locations of all of those ports. While it will be impossible to exactly replicate the connection because you have an individual throttle body intake manifold, that should be able to confirm or disprove what I have said.

 

[Adam C.]

What you have Kalun_D is a serious system dynamics situation. To envision this think of your tank (plenum) under the manifold as a spring, or in an electrical system a capacitor. The plumbing running from the tank to the runners can be considered a combination of 8 masses and dampers (or resistors and inductances). The IABP valve can also be considered a damper (resistor), which is its job.

What this means is that the system will tend to want to “vibrate” at some frequency, causing the pressure in your plenum to oscillate with each cylinder intake stroke. This oscillating pressure is not any good for your sensors or IABP.

So how do we kill the pressure oscillations? We could make the resistance of the plumbing from the plenum to the runners really high (small diameter tube). This is called an over-damped system, which won’t vibrate much. The restriction in the plumbing would have to be larger than that of the IABP valve itself. The problem with this is that the plumbing won’t fulfill one of its functions, which is to supply air to the engine during idle.

The second possibility is to increase the size of the plenum. This will make the natural frequency of the system be very low. If the system operates above this natural frequency, there will be very small pressure oscillations in the plenum. This is good, but we need to make sure that the plenum isn’t too large, or throttle response will suffer. When you flip the throttle, the plenum and plumbing will want to fill with air, and this takes time. The typical limit for this is about 1/3 the engine displacement by the way.

Another thing we have to worry about is ensuring that the cylinder actually gets some air during the intake stroke. One function of this is the previously mentioned plumbing resistance. The other parameter is the length of the plumbing itself. If the plumbing connecting the plenum and runner is too long, the intake event will be over before the low pressure wave ever meets the plenum, and so there will be little flow from the plenum to the cylinder. So we should keep the plumbing as short as possible.

So to recap:

1. The lines connecting the plenum to the intake runner should have a resistance no greater than the IAVP valve. So the diameter of the connector lines should be on the order of that of the IAVP.

2. The plenum should be large (up to 1/3 of the engine displacement).

3. The lines connecting the plenum to the runners should be kept short.

4. All sensors, vacuum taps, and the IABPV should be connected to the plenum.

So what does a production EFI system look like?

1. The plenum is large.

2. The plenum is connected directly to the primary runners. (equivalent to having short, large diameter plumbing).


3. The IAPBV, fuel pressure reference, and MAP are connected to the plenum in order to see “average” or non-pulsating pressures.

If you want more justification, we can get down and dirty with the math, but I think this covers it.

 

[Chris Duncan]

"What you have is a serious system dynamics situation"

Or in shade tree terms a can of worms?... excellent.

But I'm confused (but that's not unusual). I thought that a plenum was a chamber that the intake runners got all their air supply from, and that all the intake air runs through.

In this case I'm talking about a separate vacuum reservoir tank that only supplies some extra air for raised idle and serves to centralize a vacuum signal, although looking at the definition of plenum it does indicate "air reservoir".

The 3/8" line design is the largest possible diameter with the shortest possible length being 5" or 6". The volume of 50 cu in is also as large as possible but is only about 1/6 of engine displacement.

I'm leaning toward the small separate branch systems which are essentially just two 8 into 1 junction boxes. It is more cluttered with 16 small lines in a confined space compared to 8 with a large hidden tank, but it's easier to build and at least partially hidden under the cool air box.

The large tank would be harder to build but more aesthetic the only problem is it's somewhat of an experiment. The experts at Kinsler were skeptical of the large tank but I'm not sure I was explaining it sufficiently. The expert at TWM is at the SEMA show so I'm waiting to get his input also since he builds the system I'm sure he's confronted this problem before.

 

[Bill Bayard]

Kalun,

I have a TWM manifold on my engine(351W). The guys(Second Street Speed) who dyno tuned the engine(chassis), had no problems with it whatsoever.
I was a bit skeptical of the TWM vacuum plenum, junction box, and eight little hoses running to the throttle bodies, but apparently they are quite sufficient for my engine's needs.

BTW, I do not have an idle air control. It was deemed unnecessary. The car runs and idles just fine without it.
My Motec automaticlly leans the fuel out at idle, so that the engine maintains a nice smooth idle, thereby eliminating the need for IAC.

P.S. Second Street Speed works with Kinsler EFI literally every day(their own race car, as well as quite a few other Kinsler equipped race cars that they built and maintain). They REALLY know their stuff!!
Call Bill Hunsberger at SSS @(215) 257-3724, after 12 noon EST. He can tell you better than I(This stuff is a bit over my head!), what transpired when my car was on his dyno(re: amount of vacuum). As he has direct experience with a
TWM manifold on his dyno, not just hypotheses(Kinsler techs & Accel DFI guy), I suggest that you give him a call.





Bill

 

[Chris Duncan]

""BTW, I do not have an idle air control. It was deemed unnecessary. The car runs and idles just fine without it.
My Motec automatically leans the fuel out at idle, so that the engine maintains a nice smooth idle, thereby eliminating the need for IAC.""

I may not run an IAC if it's not necessary, that would solve this whole issue. The only reason for IAC would be for cold start really. There's no heavy loads like AC or anything.

How does your TWM setup do on cold start? Does the Motec raise the idle without an IAC during cold start?

 

[Paul Thompson 'Hooligan']

Hi Kalun

We use Motec on Roys GTD with Kinsler TBs and no IAC. The motor (347) idles great and did so with a more radical cam that was previously fitted.

Once main map is sorted - start up (warm and cold) adjustments can be made to give a great start and idle for just about any condition.

From cold - we fully depress throttle twice (batch fires inj) to wet manifold and press the start button.

From Hot, we just press the start button.

At no point do we have to press the accelerator to start - if you do, you are adding acceration enrichments/adjustments, that negate the point of having warm-up adjustments in the first place.

Just about all ECUs should be able to give a good idle condition if correctly configured.

We use only throttle position on Roys car now with the MAP sensor giving just atmospheric adjustments (minor) and the car runs great.

One thing though - spend some time getting all 8 throttle blades synchronised accurately and ALL flowing a sensible amount of air at idle - if not set correctly the engine may be idling on 6 or even 4 cylinders only....

Once set, we adjust throttles maybe twice a season - they remain set very well. (especially if you do not seperate the three piece manifolds during induction removal).

Hope this is of use.... /ubbthreads/images/graemlins/smile.gif

 

[Bill Bayard]

Kalun,

The Motec removes any hint of "cold start" issues.
Just pull the trigger, and it idles as if already warmed up.
I run a TPS and MAP sensor like Roy's car, and my car
runs great!

Instant gratification! /ubbthreads/images/graemlins/cool.gif

BTW, with the TWM throttle bodies, you may need to "tweak"
the butterflies a little, above and beyond linkage adjustment, for optimum synch.


Bill