Shoptalkforums.com

Jake is Correct....

You don't want the heaterboxes to absorb / hold heat

In the Winter / Cold conditions....

A factory Heat system, the air is directed with more pressure (Thermo / Flaps shut) thru the Heaterboxes.

Most people have their defrost / heat system "on".... so more volume is transfered thru the heater ducts / less over the motor.

Summer / Warm conditions....

The flaps / thermo are in the open position, less volume is directed thru the heaterboxes... If the heat / defrost is turned "off", the heatboxes Bleed warm air thru a small passage that is opened when the heaterducts are off.

This small passage is shut off when the heat is on.

This is why I run the Thermo / Flaps in all my motors... The things warm up better, the cooling air is directed correctly to the heads / top of the cylinders, and the heat doesn't Suck...

It's funny when I get a customer that questions why I won't build them a motor without this system, then later when they drive it for awhile, they say "You were right"... "It drives better than ever"..."My heat / Defrost system doesn't Suck anymore"...

Leaving this system off a Type I motor, is like building a really nice / expensive motor, then putting a Crummy exhuast on it.

- Joe

Right, I remember this from an ealier thread. Also I was thinking today about this, It is simple enough that VW probably had it right. If it would have been advantagous to not have that airflow 24/7, They would have delt with it. They didn't have everything right, but for 30-40 years ago they did pretty good.

Ron

As some have already pointed out, cavitation is a liquid only phenomenon. I think what people are generally referring to as "cavitation" is actually the fan blades approaching stalling conditions. I'm not too familiar with radial fans (except the ones in my VW) since my background is in ships and aircraft (i.e. axial fans). However, stall can be achieved at off-design fan rpms especially without a conditioned supply airstream (read this as a proper nacelle, bell housing, or inlet throat).

So, pure speculation here since I don't know the specifics of his test, but Wes's results may have been skewed a bit by the addition of intake bellhousing and throat. Thats the nature of experimenation though - how can you measure the volume rates and speed with out a contained measurement cross section.

Hmmm - makes me want to do a little more digging. I'll be keeping an eye on this thread for more info....

Baja Wes wrote: Bruce, I would be looking at your jetting. If you noticed this trend than you probably have a lean spot if your rev range. Before you were cruising above it, now you are cruising in it. Get a dyno check of your motor done, and get them to plot the air fuel ratio against rpm. It's amazing how much difference in temp the jetting will make.

When I built the engine, I ran it for a year and 15,000 miles on the jets I thought were right. Dyno tests with exhaust gas analyzer showed it to be not lean, but not quite where it should be, so I upped it half a jet size to where it is now and when my observations regarding the different rpms were made. Subsequent dyno tests show the CO to be a little richer than I would like it. I am certain its not lean.

zydeco wrote: So, pure speculation here since I don't know the specifics of his test, but Wes's results may have been skewed a bit by the addition of intake bellhousing and throat. Thats the nature of experimenation though - how can you measure the volume rates and speed with out a contained measurement cross section.

This is the exact reason why I used the SAA Fan Test Code as guidance on how to carry out the test without influencing the actual air flow rate. The standard gave dimensions of the intake tube, dimensions of the entry bellmouth, velocity measurement points and location, etc to ensure the test is accurate. These guidelines have been used to test thousands of fans and is the industry standard. So I don't see how you could possibly do a more accurate test. So I do not believe my results are skewed.

PS - The australian standard is based on (amoung others) the american standard - AMCA 210-74 (ASHRAE 51-75), Laboratory Methods of Testing Fans for Rating. You can buy the latest version, ASHRAE 51-1999 from here if your really keen;
http://resourcecenter.ashrae.org/store/ ... &view=item

PPS - Bruce, good to see you take care of your motor and monitor the jetting well. You problem is weird. Our speed limit is 62MPH on the freeway, my same 1916cc twin webered motor ran the hottest when geared to do 3100rpm, and the coolest when geared to do 2500rpm.

I would suggest fully mapped ignition as a good step forward to curing problems with excess heat when cruising.

Mechanical Advance only distributers cannot adjust for full throttle acceleration and also the perfect advance for light throttle cruising at mid rev range.
Even Vacuum advance dizzies can't give the correct dip in advance at peak cylinder efficiency that most engines require to get the correct advance over the rest of the rev range.
Load dependant advance is a must and mapped igition is getting pretty cheap these days. Use a throttle position sensor to map the load and although you may not increase the peak power, engine efficiency will be improved.


I believe this is most overlooked aspect of our hobby

Wes,
To clarify a little - my point about the "nature of experimentation" was that the labaratory setting often may not replicate the real world. I see this all the time in my job and sometimes fall victim to this myself. In most cases, it is impossible to get repeatable experimental condititions without putting some controls or applying some assumptions to the "real-world".

In this particular case, I'm comparing your laboratory test, with controlled and optimized (I've got a question for you about this in the next paragraph) inlet conditions, with the real world of a T1 engine fan sucking a high volume rate of air from its inlet which is all of 3 inches away from a 3 foot by 2 foot sheet metal plate. The fire wall causes the airflow to do a near 180 turn in the space of a few inches as it passes through the fan. I have no data to support any of this, but any fluid dynamisist will tell you that can't be good.

The question(s) mentioned earlier pertains to the inlet optimization - For my curiosity, was there any particular fan speed that the bellmouth and inlet were optimized too? Or were they variable geometry? What were the highest inflow rates you measured (say at a fan speed of 8 or 10k)?
Also, any theoritical or experimental info on how much power the fan is using at these higher speeds (I'm guessing 50-80 fps)? I'm sure some of this info is probably on your site.... maybe I should do a little surfing and check back here in a little while

oops - repeat post deleted

Wes,

I just looked over your site - interesting info. Did you happen to keep track of the engine RPM vs. Fan RPM before the belt tensioner mods?

Also, I know this is wishful thinking but did you measure fan/alt torque? I'm curious what power they were pulling when the belts started slipping. Maybe a toothed belt mod ala modern timing belts?

While I agree that fan slippage could be the cause of the fan stalling symptoms, it would have to be a significant amount of slippage (more than a few % per rev) to dramatically affect fan speed. I'm curious to see the crank vs fan speed if you have that info.

As a side note - sorry to hijack the thread. If this subtopic is of no interest to anyone, I can move this offline.

I understand what you mean about the effect of having the firewall right near the fan. Your right it could have an effect. In fact Gene Berg done some tests on the negative pressure in the engine bay due to the air for both the carbs and the fan having to squeeze through the vents under the rear window. He measured a noticeable negative pressure in the engine bay, which for sure would effect the fans air flow. Low pressure fans like the VW fan are sensitive to inlet pressure conditions. The solution is to use deck lid stand offs or the later vented deck lid. So no I don't know how having the firewall next to the fan effects it. I expect you'd have to be pretty high in the rpm range to start to see an effect.

The inlet to the fan and the bellmouth were size as to not create a negative pressure at the fan inlet, or at least not restrict the flow in any way. The air velocity at 10,000 fan rpm was over 200kph! (125MPH).

As for the fan belt slipping, no I didn't keep a log of it. But I can tell you that with the standard fan belt set-up we could not get the fan to reach 10,000rpm. It got to a certain rpm and just would not go any faster as the belt could not transmit any more HP to turn it faster. It was slipping so much we got powered rubber built up on the alternator standard. We completely wore out a fan belt. Then we built a tensioner to bend the belt back under the pulleys further and increase the angle of belt wrap on the pulleys. This let us get to 10,000rpm, but we couldn't get it much quicker than that. We could have the engine doing 3,500rpm and the fan would do 10,000rpm, then we would increase the engine to 5,000rpm and the fan would stay at 10,000rpm and rubber would start coming off the belt. It was like smoking the tyres, but smoking the fan belt instead. A modern multirib belt would be better at transmitting the power.

Unfortunately we didn't measure the fan power, althought it would've been great to do it. I expect at higher fan speeds it's quite alot as the VW fan belt couldn't transmit it. Because of the high power required at higher fan speeds, it's probably better to have the belt slip so you don't loss so much HP to the fan.

It is pretty simple to work out how much Hp the belt can handle. If your not an engineer you can always look in the back of a WW Grainger catalogue. They have charts for belt width, length, and pully angle = HP.

Remember the BB serpentine belts? They are not on his big adds anymore. Actually I think the serpentine belt was CBs Anyway, I think they used timing belts, which can handle more power.

Ron

Bruce...well put...but we are quibbling over my late night lack of vocabulary. By what rate of advance...I mean...excatly how many degrees of advance are you advanced to while running at constant speed? Because you will be advanced and you may run hot from that if advanced too far.

So...its not a power pulley right? cause that would be smaller than stock. But then...a 12% larger pulley...would drop the rpm slightly...correct? Maybe you are running the fan slightly slower at speed...than the cooling system requires? Its something to explore.

Cavitation. The term is for fluid.....I realize that. Denotes shearing of gas from liquid. The effect in a fan from turbulence, is much the same in what it does to flow...and exists just the same. All fans have stall speeds.

Mharney...the last part of your post is what I was getting at. You do not have to be sonic in fan blade velocity, to create turbulence in a fan, great enough to cause partial blockage or resistance. You said it better than I.

Baja wes. Yes...your fan calculations are common in most HVAC books. One thing you are missing is that those do not measure the "stacking" of all elements involved in any given fan installation.
yes Rpm and output are linear....or can be...but only in unducted form and with no restrictions...ie: head pressure, or supply duct being smaller in cross section than the max rpm flow rate of the fan...requires. The calculations show none of that.
Use that fan to start pushing air into the cooling manifold of a VW engine...or an AC duct with a large volume of air ahead of the fan....and numerous constricted duct openings (or any other number of examples)....and you will reach a point where the resistance on the output side of the fan...creates turbulence at the tips of the vanes.
Centrifugal fans...like these..cannot tolerate that. Flow is laminar at the tips. Air in these fans is not pushed by vanes..it is flung from the tips due to centirfugal force.
The design velocity at the vane tips, the area of the vanes, the shape of the vanes, the design HP allowed, and the design rpm....are also largely dictated by the type of centrifugal (or radial ...whatever you want to call them) fan in use. There are 3 basic types....forward canted, backward canted, and one that can only be described as composite with supercritical helical blades....and can only be properly used in a duct.
It is that "flinging" movement, noted above, that creates velocity....that creates the laminar flow at the vane tips....that continues the process by creating a depression on the leading edge and underside of the vane...drawing in air. The design, texture and angle of the fan floor base also contributes greatly.
Turbulence from head pressure...which builds as rpm rises....can cause a breakdown of laminar flow at the vane tips. It appears almost as "slippage" and many people have described it that way. Which is why I term it cavitation (though it is not). Its effects are very similar. Efficiency of the fan at that point DROPS...though the rpm may continue to rise. The math doesn't quite describe that (although there are formulas to describe the effect...when calculating head pressure and frictional loss).
The only way to solve it...is to vent head pressure. This is a totally different effect than what happens when a turbine goes "sonic". In that scenario, heat and vapor expansion create localized "sonic booms" or violent charge expansion, that may actually reverse flow on a turbine with catastrophic results.
Just as a thought, the belt slippage you note...is not prevalent in type 4's like I drive. They are crank driven. In the stock system...by my experience, limiting turbulence begins around 4700 rpm. I find reductions of about 20% by 5200 rpm (which was all I could get out of it ). It was measured as reduction of velocicty at a preset drop rate of 3% by a hot wire type anemometer. That simply means that I pre-set it to squeel when the rate started going negative instead of positive.

Though I appreciate and undertand the formulas you provided....I propose that they alone, are not enough to explain all the aspects of the application we are speaking of. Fan volume cannot remain linear with rpm in a restricted system. Sooner or later something has to give, In a type 1...as you note...its the belt. In a type 4, it turbulence...loss of cooling air...and increased braking action by load, on the engine.
As I have just started to discover in the last year and a half, a smoke cartridge and a camera can greatly distort the results that the math swears we should get. I havn't started playing with VW fans this way...but its been enlightening on the manifold. It has been enlightening in the past...in HVAC centrifugal fan systems...where I first used smoke to define fan stall due to excessive head pressure. It was happening where the math swears it shouldn't.

Just a friendly note please....don't poke at peoples education too hard. You might be stunned at what you find in this forum. My poor typing may not allude to either one of my degrees, my years of experience in process controls and systems or my position on the process engineering staff in a semiconductor facility. Aside from all of that we can only be familiar with what we use. The rest of it we eventually lose through dis-use. Ray

raygreenwood wrote:Bruce...well put...but quibbling over my late night lack of vocabulary. By what rate of advance...I mean...excatly how many degrees of advance are you advanced to while running at constant speed? Because you will be advanced and you may run hot from that if advanced too far.

So...its not a power pulley right? cause that would be smaller than stock. But then...a 12% larger pulley...would drop the rpm slightly...correct? Maybe you are running the fan slightly slower at speed...than the cooling system requires? Its something to explore.

The timing before and after the tire change was the same. I cruise at 3500-3700 rpm and the advance is all in by at least 3000 rpm, so I don't contribute the increased temp to either the timing or jets.

The 36hp lower pulley is about 12% larger than stock. Since its on the bottom, it will speed up the fan by 12% (assuming no slippage).

Ah....yes! Thanks bruce. I was thinking backward. I'm used to crank driven fans (Doh! like the generator would be driving the fan...and not vica-versa ).
12% doesn't sound like too much...at 3500 rpm, the fan would be turning 420 rpm faster than normal. I can't remember what the normal ratio of crank to fan rpm is.

Just for giggle and grins...open a controlled vent in the sheet metal...say on the back corner areas farthest from the fan. If the problems lesssens...but only at higher cruising rpms....then you have what I was describing. Cooling air stacking up and head pressure too high. In other words...the fan being too efficent for the constricted volume it is supplying. Venting it would most likely be a problem at the lower rpms though, as at that point the vent would be cooling air loss. Just some thoughts. Ray

Ray is someone who is wel respected by many here, including myself. I have never seen him be one ounce out of line with his recommendations and experience. He has proven to be a huge asset to the expansion of accurate knowledge of the forums.

I'm not an engineer or a lab technician... But I handle more of this stuff in 3 months than most people get to in a lifetime.