Xfsaer wants to build a fan car type FSAE car for autocross type demonstrations

[Xfsae]

Attachment 90

Cornell 90 sucker car let's reverse engineer it

Suction area doesn't seem to extend under the engine
Hope i had these image clean up skills to have a better look
What do you guys and gals make out of it?

Also how does the suction work?I see no fans?
cant see fans parallel to the floor or ducting leading to a remotely placed fan .Car seems pretty clean as a design
BUT i see this aluminium hose leading to the floor ( look low in front of the rear wheel) .Could it be there are 2 vacuum hoses one on each side leading to the vacuum box in the floor?
Could a standard vacuum pump or 2 as used for refrigeration work ? I mean for a Cornell type solution and not an all out sucker car as Z described above (like for a first attempt attempting to get the benefits of big wings seen today in FSAE from the sucker layout - first target- max 1 ton of downforce and no more than 2.5-3 gs lateral )

Is this Cornell guy blipping the throttle or flipping a switch to feel the downforce in a standstill just before the start of the skidpad?

[Z]

On the issue of "how much performance gain is possible from active-aero (= "vac-trac")?", there is no doubt at all in my mind. There is, in fact, SO MUCH PERFORMANCE GAIN possible, that everyone and anyone that has ever come near it finds it quite terrifying.

It is instructive to read the history of the Chaparral 2J. Very briefly, (IIRC) Jim Hall was recovering from a major racing accident during the 2J's development, so he couldn't provide much motivation. For the most part, only one engineer and one mechanic worked on it, because the rest of the Chaparral team hated it (it was the unloved, ugly-duckling, "skid-pad racer"). It was entered in four races, but never won any, because of various little teething problems and its lack of resources (only the two guys...).

But its performance in the four races was enough to terrify the whole of the supposedly "anything goes, let's really push the envelope" Can-Am field. It was banned by unanimous pressure from the whole field, except for Hall himself. In fact, the banning of the 2J was the tipping point that pushed Hall out of motorsport.

Personally, I see the "fear" of vac-trac coming mainly from its apparent open-endedness. Where is the limit? Is there a limit? Is the limit when the driver's body, or what is left of it, has to be drained out of the bottom of the cockpit via a sump-plug? Essentially, IMO, the potential limit of vac-trac cars is so far in excess of what any driver could take (or would enjoy taking!) that very few people want to play with this particular toy.

Because of above, I reckon the future of vac-trac is limited to demonstration type events. And because the potential horizontal G forces are so hard on drivers, the demo-cars may have to be remotely controlled (ie. "play-station" drivers, tele-connected to the cars).

However, here is an easy-on-the-driver alternative. During a break in the 800+ hp "real racecar" racing, a small, lightweight, perhaps go-kart sized, car is rolled out in front of the grandstands. Four men grab a corner each and pick the car up to prove that it really is lightweight. A stout rope is tied to the rear of said small car, with the other end tied to the rear of one of the real racecars (V8 Supercar, Nascar, F1, whatever...). The two cars start their engines, and ... engage in a tug-of-war.

The real racecar has smoke billowing from its rear tyres, but it is going backward! The little vac-trac car has no tyre spin, and only a small engine, but, in an appropriately low gear, pulls the screaming monster the length of the main straight...

Well, I would find that entertaining!
~~~~o0o~~~~

Xfsae,

Regarding the "45 hp" engine to power the fan, that figure was for the Chaparral 2J, which had a 700+ hp (?) main engine (8-litre V8), and lots of skirt length. For an FSAE-sized project I figure you would need less than ~10 hp. The power requirement is directly proportional to the amount of skirt leakage, and the less of this, the less power you need.

I suggest the following development path. Take four 1 metre lengths of 50 mm x 100 mm (2" by 4") pinewood, and join as a 1 m square "picture frame" placed on a level concrete floor (frame is 100mm high). Take a 1 metre square by 10 mm thick sheet of plywood, and glue&screw to the top of the frame. Drill a ~30 mm hole in the plywood, duct-tape your mum's vacuum-cleaner to the hole, and switch on.

This vac-box will quite obviously have downforce. Seal the floor-to-box edges with some duct-tape (or plasticine, blue-tack, etc.), and the downforce will be more obvious (the plywood roof will bow further downward). So time for some measurements.

Screw some more lengths of pinewood to the top of the box so that they extend outwards and can be placed on top of some weight scales (bathroom, or professional racing ones...). Try different levels of skirt clearance (1mm - 10mm), and also with the skirt-gap completely sealed with a flexible membrane (eg. duct-tape). Based on my earlier tests your fully sealed vac-box + vacuum-cleaner should have about a ton of downforce, but less when there is significant leakage.

So time to get a bigger fan! My quick googling (for ~ "industrial high pressure blowers") found a smallish, moulded plastic, centrifugal fan, about 0.3 m outside-diameter and about 0.1 m in axial length. The pressure (P) vs flow-rate (Q) diagram for this fan had a smoothish curve from P = 20 kPa (= 2 tons/sq.m) at Q = 0.056 m^3/s (almost stalled), to P = 10 kPa (= 1 t/sq.m) at Q = 0.25 m^3/s (near max flow). Pressure actually dropped slightly at stall, and these numbers are for "suction".

Power required is simply = P x Q x Efficiency, so given that efficiency (while not stalled!) is usually above 50%, the above fan can be powered with a ~5 kW motor (as was claimed). So get a ~10 hp industrial IC engine (Chinese copies of Hondas about ~$500 here in Oz), or borrow from your lawn-mower, and arrange appropriate belt drives, suction-ducts to the vac-box, etc., then re-test for downforce vs skirt gap...

Depending on skirt arrangement (which could be a lot more development, though fairly easy if done as above), I reckon the above vac-box should easily see ~1 ton of downforce. Or maybe 2 ton? Anyhow, bolt under your FSAE-sized car, and enjoy!

Z

[Xfsae]

@ RJWoods77 reading through the cheaparral website now - thank you for the link
They used a fan from M1 tank ...It is one of the intake fans at the back of the tank

i was trying to see if there was a specific reason for them choosing this fan-12000cfm at 70 pounds weight must be the reason ... It is a 2 stage vane axial design in alloy1-2.JPG

Some useful info from the corvette "cheaparral" fan car project:


"Extensive calculations and testing with a proto sled identified the ammount of airflow and vacuum required to stick the car to the ground with 1500 pounds of force (6675 Newton or like 680kg extra) ,the ammount necessary to boost the Vette's rating from 0.9 to 1.4g lateral "

I wonder how they found that they had to add this ammount of downforce to get this specific increase in lateral acceleration given that the ultra low budget of the project should not allow them to buy tire data - not even new tires- Did they start loading the car with weight and driving around with an accelerometer ? Any idea or suggestion?

The Abraams intake blower seems to be about 0.4 m in diameter and 0.4m in height judging from the photos of it placed in the Vette and on the tank and was powered by a 33hp snowmobile engine in the cheaparral project

4.JPG (placed in the codriver seat place on the Vette)

At 10 inches of water operational vacuum (that is -14.33 psig relative to atmosphere ) they got 1000 pounds (454kg) of downforce
They measured it with a vacuum meter on the car - not sure how they were deriving the downforce amount

To maintain traction throughout the suspension travel a 2 piece skirt frame connected by bellows was developed and installed under the car - shopping trolley type of caster wheels attached to the lower frame would maintain a 1/2 inch gap between the skirt and the ground

untitled.JPG (skirt- trolley wheel photo)

Z- as always interesting , informative (and entertaining in a cool manner in this case!) post .Thank you

I was looking in the internet for performance curves ,dimensions and weights for plastic industrial high pressure blowers of the dimension you suggested

http://www.mkplastics.com/documents/...MARCH_2008.pdf
http://www.plasticair.com/dat/files/28.pdf

is what i come up with but they dont seem very promising (in terms of space, weight and power required) ,in comparison to the cheaparal project figures i found and mention above

can you post a link to the high pressure blower you found?

George Bowland.JPG

FSAE fan car boy-George Bowland's leaf blower vacuum assisted B8R A-Mod

leaf blower said to have been running WOT

Skirts seem to be of polyethylene or nylon variety -someone mentions them as rubber skirts but i dont know they dont seem like that to me- they seem to be actually touching the ground on the sides - i don't know how they seal front and rear and in the corners

[rjwoods77]

XFSAE,

Forum etiquette would have you edit a post with addendum vs creating a new post for each addendum. You can do this by clicking on the "Edit Post" button. Consolidating your thoughts before you write as well as combining topical information will keep us all from getting an eye beating.

[Francis Gagné]

XFsae,

First, I have the same comment as rjwoods. Please do not double post, it is not a higly regarder behavior on any forum that I know of. (Plus this thread really ought to be in the Off-Topic section)


Back on topic,

I have not read much about the Cheaparral, so I have not verified your input values. But if we do the maths :

10 inH2O = 0.36 Psi = 2500 N/m^2 (not 14.33 Pascal as you have mentionned) so nowhere near full vaccuum!

They say it pulled 1000 lbf, so about 4500 N. Assuming constant vaccuum under the skirt, we would arrive at 4500 N / 2500 N/m^2 = 1.8 m^2 of skirt area. Which is plausible.

Using the 680kg supplied in your quote, to obtain the 0.9g to 1.4g they would have made their calculations using a curb weight of 1225kg. The C4 Corvette curb weight (from Wikiped) is 1469 Kg, so again plausible. assuming the stock car goes 0.9g around a skidpad (the 0.9g as probably been measured on a skidpad with a clockwatch). Assuming all else equal, adding 460kg of downforce would results in increas of (1225 + 460) / 1225 = 1.38x This means the car would now pull 0.9g x 1.38 = 1.25g

[Xfsae]

@ Francis - Hello and thanks for stopping by to contribute

= As far as i can see, i never double posted (posted the same thing 2 times)

about this being in the off topic thread - i dont see it being a huge deal (given the number of views we are getting here i dont think people care that much) but anyway if the administrator here say it should be moved , they can do it (at least in other forums i have been to they seem to be moving threads without too much fuss all the time)

http://www.cheaparral.com/
(in the first page and the links to the documents in the right are all the info)

Never mentioned pascals anywhere
I headed here - in the second category that is about vacuum:

http://www.convertunits.com/from/inc...column/to/psig

There seems to be a general confusion regarding psi vs psig but this convertunits.com seems right-according to them 10inches of water column (relative to vacuum)= -14.33 psig (relative to atmosphere)

Also i am aware that 10inH20 =0.36 psi if you are measuring differential pressure, such as the difference in psi between two points. It also gives the correct answer for absolute pressure, assuming you are measuring psia, which is the pressure relative to absolute zero vacuum.

But if you are measuring relative to vacuum and want to resolve the pressure relative to the atmosphere, then:

10inches of water column (relative to vacuum)= -14.33 psig (relative to atmosphere)

so i dont agree with you Francis that they were nowhere near full vacuum ( full vacuum= -14.7psi)

again you can use this for less confusion:

http://www.convertunits.com/from/inc...column/to/psig

Anyway , in relative difference terms your calculation still stands - pressure difference they achieved=10 inH2O = 0.36 Psi = 2500 N/m^2

Given the mentioned 1000 lbf (or apx 4500 N) of downforce achieved & assuming constant vaccuum under the skirt, we would arrive at 4500 N / 2500 N/m^2 = 1.8 m^2 of skirt area. Which i agree is plausible and the ideal skirt dimensions for this type of pressure difference would be those IF there are no leakages ( vacuum box sealing perfectly to the floor)

In fact i found today this spreadsheet in the 3rd pdf in their website GRM-3 :Estimated fan requirements.JPG

From which i see :

target = 1000 lbf of downforce
Skirt size = 48in x 115in =5520 sq in (or 3.48m2)
By leakage under skirt they mean the distance from skirt to road =0.5in (as i mentioned above)
Calculate pressure required to achieve the target downforce
= 1000lbf/5520sqin= 0.1812 psi (5in H20)
(no consideration about leakages)
In another article as i wrote above they mentioned that they actually needed DOUBLE that ammount of pressure difference (10in H20) to achieve that ammount of downforce in practice
i guess this is due to leakages mainly




@ rjwoods77 - the thing that i can make out from your comment is that i should try to put in the same post the info of mine?there is some time difference between my posts always ( i dont deliberately break up my posts ) - someone else might post in the meanwhile and he/she might not have noted i changed something above - i personally prefer to see posts separate (as done in a messenger twitter type situ) so that i know what was posted and when in retrospect

I moved some consecutive posts to the same one- hope its better for you (dont know why most of the links to my pics were not working though after some hours-initially i checked them and they were all working) - for me it was not visibly disturbing to see the small blue bar between posts but i did for you guys and gals to enjoy it Hope its more pleasant for you now

Also excuse me for my ignorance but if i edit a post by adding info ,how does a third party reader know ? how is he/she notified ?- i cant believe they will easily remember the last edited date after some weeks). Also does editing a post make the thread surface ? ( i mean show i some way that there is activity in this thread comparable to making a new post so that people understand there is movement there?)

For example if i want to ask a question at the end of a post that i have done before that did not include a question , how is somebody else to know if i consolidate it with the older post if he/she is not just regularly opening the thread lets say out of their own curiosity? Thanks

[Z]

Xfsae,

As exFSAE said a few pages ago, your posts do come across as being "way too scattered". And as Rob said, you ARE repeating yourself. Quite often. To keep people interested in the subject it would help if you prepared your thoughts a bit more clearly before posting them.

Francis's numbers above are in the ballpark. I guess the Cheaparral crew managed to lower the mass of the car a bit more (just stripping out stuff), and hence got a somewhat higher predicted cornering G for their 1000 lbs vac-downforce.

Based on the link to Cheaparral's "Estimated fan requirements" (some funny/wrong numbers there!) I would strongly suggest ditching imperial units (inches, pounds, psi, etc.), and stick to the following units, mainly because these numbers are so easy to remember and work with.

1 x "atmospheric pressure" is (approximately) equal to;
1 bar,
1,000 milli-bar,
1 kilogram (taken as a force!) per square centimetre,
10 Newtons per square centimetre,
(1 ton per square foot <- ok, this imperial unit is easy...),
10 tons (force!) per square metre,
100 kilo-Newtons per square metre (100 kN/m^2),
100 kilo-Pascals (100 kPa),
10 metre high column of water (which, if cross section = 1 metre square, would weigh 10 tons).

So, from above, if you want 1 ton/sq.metre downforce, then you have to lower the atmospheric pressure under the car by 1/10th (= -10%), which means finding a fan that gives a pressure difference of ~40 inches of water (because 40 inches = ~1 metre = 1/10th of that 10 metre column of water).

Here is a fan that comfortably delivers this suction (and twice as much...) from a motor of less than 10 hp.

http://www.aircontrolindustries.com/.../01._ep10a.pdf

Note that this is a "backward curved blades" radial-fan of aluminium construction (I can't find the plastic one?), and most of the quoted weight would be in the three-phase electric motor. Also, I would not trust their quoted RPMs (20,000 rpm!, I reckon they must get the tea-lady to prepare these data sheets!!! ).

Z

(PS. The image of the Cheaparral's vac-box appears quite a bit smaller than in some of the above calcs (eg. the "fan" ones). Hence, I guess, their move from 5" to 10" w.g. to get the required downforce.)

[Xfsae]

Hi Z and thanks for the comments i will go through your latest info as soon as i can .I agree 20000RPM from this set up doesnt seem likely - more like 1500 to 3000 RPM max... 20000RPM only if it was a super precisely balanced brushless RC ducted fan for giant scale RC aircraft

And people loosen up a bit please - i understand that you are tired of people asking "can you tell me how to design a suspension" or similar but you see that this i not the case here - i do my research, take the time to share it for all the people that might not have the time to go search for individual resources and I read through all the input i get here as carefully as i can despite this being not a thesis project for me or similar , so please people dont bite ... Want to make clear once again that i am doing this for fun - i dont have to write an essay or report here

I am running through the cheapparal numbers so would help me to check together the thought process behind them:

- I get a fan flow rate of 166,98 m3/s (or 170m3/sec as they mention in the spread sheet) through a combination of the Bernoulli and continuity equations if i use as suction area the sum of:

plan area of the skirt suction area (top view area of the vac box) =5520 sqin

plus the leakage area (the sum of the 4 rectangular pieces of gap formed between the bottom of the skirt and the road :

2 x 115 x 0.5 + 2 x 48 x 0.5= 163 sqin

Do you think that this is correct as strategy? What happens if for example we use skirts touching the road ? we just ignore the leakage area? - doesnt seem very correct to me ...What are your thoughts on that?

- Fan flow rate of 166,98 m3/s (or 170m3/sec as they mention in the spread sheet) is not equal to 10192cfm but 360209,6 cfm wherever i checked - i think that is a mistake on the spreadsheet - if there is something else happening let me know

- Bernoulli equation is used for incompressible flows mainly - do you think that it is good practice to use it here for the twin stage Abraams fan ? I mean if it was a slow rotating non ducted propeller/fan we can make those assumptions ...but here? I understand that there must be a simple method to get the ballpark cfm but is this the best?

I want to make a similar spreadsheet for baseline calculations so any input on my questions/remarks above is welcome

Z yes you are right - the dimensions of the vac box in the spreadsheet are definitely not the ones of it in the finished car .Judging from the comments on the spreadsheet they must have been going for the max possible area (as is reasonable to rip max benefits) but probably when they started designing in detail the (4) panhard rod type system that allows the skirt to move with the suspension travel they saw that the kinematics of it would be difficult to achieve in practice eg when the car is pitching a lot. The Corvette C4 wheelbase is 2440mm and in thisside car view connected 1.JPG side view the length of the box seems to be apx half the wheelbase (1200mm)

Also in this photowidth of skirt more or less equal to track width.JPG the width of the skirt seems to be about equal to the track width (1500mm). This results in a 1.8 sqmt area ,apx half that mentioned on the spreadsheet (3.5 m2) and due to the linearity of the equations above , must need twice the pressure to give the same downforce

[exFSAE]

I still feel like this would be best served in a different sub forum, or better yet in a blog.

Still feels like this is very much "stream of consciousness" style or scattered writing.

[Xfsae]

@ Z - looking at the last unit you posted above , if the 7.5 kw motor is anything like that:
http://www.motorcontrolwarehouse.co....prod_1213.html

it would weight around 40 kg (out of the 70kg of the complete unit)

if the motor is replaced with an industrial type diesel engine like that:

http://www.amazon.com/10-Air-Cooled-.../dp/B008FQU6T6

which weights approximately 50 kg (shipping weight is mentioned as 120 pounds)
and add sealed floor, skirts, ducts and other arrangements we get to more than 100kg for the suction system ( 1/10th of the required downforce is just found and it works from a standstill )

There is the option of connecting to a smaller engine but what about its efficiency/performance?
I am looking at the curves on the unit Z posted above

I guess the 2 separate diagrams are for the vacuum exit/entryvacuum performance curve.JPG and the pressure exit/entrypressure performance curve.JPG of the unit .
Typical textbook information suggest that a centrifugal fan performance curve should look something like this :Typical centrifugal fan performance curve 1.JPG

Looking at Zs attached document i can identify the static pressure curves in bold black but not the horsepower curve(s) - we need at least one Static Pressure curve and on Horse power curve (for the same rpm) from which we can determine power requirements and efficiency at our decided operating point. What are inclined dotted lines representing ?

ATM looking for a good book/source on radial fan design as there are several design variations of them but to be honest i am in deep thoughts about why 2J, BT46B & Cheapparal all went for axial flow fans.Could it be that in practical terms the seal of the vacuum box is never close to perfect so you have to move a large ammount of air/min (CFM rating high) to achieve the pressure difference needed ,something that could possibly lead to a big and heavy centrifugal fan of conventional design?
I know that there are several types of centrifugal fan variations, but Z the CFM values mentioned in the diagrams of the unit you attached above are low compared to what i should look for if i follow the thought process of cheaparal (as i described and asked for people her to double check in my post in the previous page)

This2.JPG is the vacuum test rig of the cheaparal project the size of the centrifugal fan & the 3phase motor (which seems pretty hefty -30hp?) they used to get the pressure difference necessary leads me to the same thoughts.The vacuum box on the test rig seems very close to the one used for the finished car

I am in thoughts also about what was happening to the BT46B at low speed corners(did they have so much downforce that they didnt care or the RPM variation was not big enough to vary downforce hugely?) ? The fan was connected directly to the gearbox main shaft via a single gearing stage , if there is no other trick the downforce must have varied widely according to the level of throttle. Also everywhere i read it they say that they used clutches to prevent over reving of the fan (and possible destruction due to centrifugal forces) during gear changes. What kind of clutch arrangement could that have been ? something directly opposite in principle with a centrifugal clutch? Or just a smaller conventional plate clutch also connected to the main clutch actuation ? What do you think?10.jpgThere is also the gearing from the main gearbox shaft to the fan which was calculated by an expert alongside with fan design-given that this was not a variable pitch fan - this was optimised for a single operating point - i wonder how they defined this point...
Though i cant find official figures of the levels of downforce achieved , i did dig out that they had to change from 1000lbs to 3000lbs springs from BT46 to BT46B !!!! (same suspension design for both- no rockers or anything similar introduced between the versions , just conventional outboard springs and dampers at the back and the same pullrod setup in front).Ermanno Cuoghi, Lauda's chief mechanic said that they had to do that otherwise the car bottomed out to its belly when revving up with neutral!!!

What do you people think?