Anyone used a VH44 as a clutch booster?

[Herbmiester 923]

25% accurate at the very best I would say or as I see it 75% inaccurate, yet you post it up misleading people with fast and loose calculations.  The facts as you present them, are as much opinion and support for your way of thinking as anything else. You really do waste peoples time with this type of approach. 

[M3AN 4016]

I'll take it out of the mathematical realm because, to be honest, I'm not following the math.

Could @3pedals numbers be considered representatively accurate even if they're not literally accurate? i.e. could they represent the real-world outcome even if they're not precise? Genuine question, I really don't know.

Either way, I do agree that empirical, experience based evidence is more valuable than theoretical modelling. 

Edited June 10, 2020 by M3AN

[Herbmiester 923]

The reality is that no one will do any form of testing other than drive the car and see how it feels, hence why using proven systems is the best method. Basic modelling is close to worthless when it fails to to compare with what is actually required, as an example, how much torsional stiffness is required to give a suitable solution? I dont know, and I suspect you dont either, is the UUC flywheel stiff enough? It would seem so and similar Alloy flywheels used in racing applications for decades would seem to support that.   And again because of the very basic calculations and assumptions made the basic modelling is nothing more than calculated guesswork, borderline worthless. Unless you were doing some Finite element analysis then the basic work you are doing is irrelevant compared to evidenced based feedback. 

[B.M.W Ltd 950]

My clutch setup was still running fine when I sold the car in 2018

[Herbmiester 923]

To be clear I never suggested parallel testing, that was your idea.  The reason I did not suggest it is that apart from what you feel it is essentially worthless without some form of laboratory to monitor results. Racing is different from the street but it puts more strain on the components and as to maintenance that  is to a certain extent irrelevant in the context of this conversation. Stating that alloy flywheels have a habit of exploding in race cars is a very broad statement and without context again, irrelevant. Sating that torsional stiffness is a byproduct of material and geometry is obvious to everyone but without  reference to usage and at what levels it will fail, deform etc again its an irrelevant statement. I am not trolling you I am making you account for broad sweeping statements and loose and fast calculations that you use to justify your position.  The fact you only refer to your setup as Combo X is  also interesting as we dont know what combo X is no one here could dispute your numbers. So no not trolling you just making sure that people who come here looking for information do not get led down the garden path by you. I say this as you have a history on this forum of extreme bias towards anything you champion, and that is a factual statement. By the way I am not trying to justify my position, I dont need to as my position is simple I have something that is know to work.

Edited June 11, 2020 by Herbmiester

[tawa 150]

On 6/10/2020 at 10:41 AM, 3pedals said:

Error in summing the Fid -M5 cover mass levers , also carried into torsional stiffness - revised table attached , have also included Fid with M3 cover. 

 

In response to rotational inertia vs Kinetic energy - the first is essentially treated as a static lever - mass at distance,  the second is a velocity squared based calculation the longer lever with less mass of the UUC produces  higher velocity at 5,000 RPM  the product is higher. 

Agree with Polley -   match the pressure plate to the engine characteristics - you also get some benefits as per above 

Also agree with Glenn - have fun and part of my fun is crash testing the so called proven solutions - the lighter stock  or M3 pressure plates will work  fine  with largely stock or lightly / moderate engine mods. 

 

 

 

Velocity for rotational inertia is given in radians per second, it's only the rotational inertia term which changes the result for kinetic energy so it's a simple scaler with rotational inertia isn't it? What equations are you using?

In other news, the VH44 turned up today, might see about positioning it and getting the right fittings/lines this weekend...

[tawa 150]

19 hours ago, M3AN said:

I'll take it out of the mathematical realm because, to be honest, I'm not following the math.

Could @3pedals numbers be considered representatively accurate even if they're not literally accurate? i.e. could they represent the real-world outcome even if they're not precise? Genuine question, I really don't know.

Either way, I do agree that empirical, experience based evidence is more valuable than theoretical modelling. 

I've got some questions regarding some of the calculations, but the core idea is certainly good, rotational inertia can be estimated across different flywheels easily enough to get a relative result, which ones are 'heavier' than others and take longer to spin up/down.

[B.M.W Ltd 950]

What parts/type/brand are you using in your Combo X example Ron ? and is this the combo you will use in your vehicle?

[KwS 2425]

On 6/7/2020 at 8:02 PM, 3pedals said:

Glenn outlined the difficulties in terms of the possible location and queried the "operation"  

Boosters work fine in brake circuits because adding  'blunt instrument'  technology doesn't detract too much because position is not critical, only the pressure.

However doing the same in a  clutch circuit where feed back and absolute position are critical may (will most likely)  result in a very ordinary outcome. 

 

 

Back to this; were you aware that Mitsubishi actually used a booster on the clutch on the Twin Turbo GTO from factory? Its a far more complex setup than the VH44 one, being under the dash and all, but same concept.

There is a lot of good discussion here about using the VH44 booster on a clutch, https://www.mr2oc.com/threads/lighter-clutch-via-vh44-booster.456318/

[B.M.W Ltd 950]

We used the same boosters back in the late 80's early 90's to replace 2002 brake boosters. The system worked well and wasn't that hard to do.

Edited June 12, 2020 by B.M.W Ltd

[tawa 150]

9 hours ago, 3pedals said:

Missed this one earlier - as commented earlier rotational inertia does not take velocity into account because its a change of state and the state can be "at rest" in which case there is no velocity. 

As per the example I gave earlier : No,   Kinetic energy is  not a scaler it is a square of the velocity and the velocity as outlined is linked to the lever arm.  

The formulas are the standard physics ones taught at Scholarship level physics. 

P.S really surprised you are pursuing the VH44 option - changing the slave cylinder is much easier   and likely to give a better result BMW did it with the M3 and many others. 

The velocity is in radians per second, it not linked to the 'lever arm' at all, I think your table and interpretations needs some revision. But of course I'm open to a counter-argument if you have a formula which states otherwise...

https://courses.lumenlearning.com/boundless-physics/chapter/rotational-kinetic-energy/

As mentioned in earlier posts, I like the pedal travel/engagement where it is, just not the force required.

[tawa 150]

28 minutes ago, 3pedals said:

The Flywheel is out of Europe  and comes with an 'optimised' (their words) E39 530d pressure plate . What is different about  the flywheel it is that it has more meat in the middle (thicker and no lightening holes) but thinner up behind the friction  plate where the strength is provided primarily by the steel friction plate and the weight is not required / desired. 

The pressure plate is supposed to be similar in weight to the OE 328 one but good for  380-400Nm 

The sprung friction plate I'm not sure yet as I am keen to use the M5 one but am having trouble getting one. 

 I've got no idea how long it will take to get here but when it does I'll be doing some measurements and calcs prior to install,

Once I have decicided it meets the requirements I'll detail it then install and report back - If it doesn't cut the mustard it will be a large paper weight and its on to plan B 

then when I have the OE dual mass out I'll be able to measure that as well. 

 I've got an idea to use my battery drill to try and spin the flywheel and cover up from stationary

The methodology is set the torque on the battery drill to the lowest and  increment the torque then record the setting when the flywheel moves without activating the torque clutch on the drill - I can them measure the torque with a dial type torque meter. 

 I can then repeat the exercise on the other fly wheel and I will have a relative "moment of inertia" number 

I haven't worked out how to do the Kinetic energy measurement yet but I have a couple of ideas and time . 

Prony brake would be a far better option imo, easy to make and perfect for pretty accurate relative measurements.

[tawa 150]

13 hours ago, 3pedals said:

That's the formula for  kinetic energy (the second part) and what I used and  yes you can use radians if you wish.  

You are missing the first part of the discussion  which is inertial rotation the lever arm comment refers to the first part 

It's not so much 'use radians if you wish' as it is use radians because they are the correct term and unit for the equation.

And it's your understanding of kinetic energy that is flawed, so it's that part of the discussion I'm having.

This comment:

On 6/9/2020 at 12:58 PM, 3pedals said:

UUC combos have lower inertia which means it will spin up quicker however once spun up it has higher kinetic energy,

Violates the law of energy conservation, if it spins up quick, it takes less energy to spin up, so has less kinetic energy.

And this table:

On 6/9/2020 at 12:58 PM, 3pedals said:

Shows a misapplication of the kinetic energy formula, for a constant angular velocity (as you specify), the kenetic energy will be linearly proportional to (scales with) the rotational inertia.

 

[tawa 150]

2 hours ago, 3pedals said:

Hey , good to have a bit of fun with this.

You are correct with the comment on angular velocity  if I were using a constant angular velocity, but that is not the case,  and the key indicator to this is my  repeated comment about the mass location.  

Constant rotational frequency does not translate in to constant angular velocity.  

·         The purpose for using a constant rotational frequency (5,000RPM) in the calculation is so that it cancels out and does not influence the results.  

·         The radius at which the mass is centralised determines the angular velocity,

·         The various mass loci occur at different distances from the centre so they have different angular velocities at the same rotational frequency and that is what is used to produce the numbers in the last column.

 If we use the linear equation   (½ C.V²) this is easy to demonstrate: Imagine I am a  hammer thrower and I have a 10kg hammer on a rope.

1.       I make a  2 metre circle above my head ( 1m radius) and release the hammer at 10 m/S  . At the moment of release the angular velocity and the ‘velocity’ are the same so we can use the linear formula  at that point. For the hammer travelling in a more or less straight line we get an answer of: 500 ( (10*100)/2)

2.       For my second attempt I hold the rope a bit further down and make a 1.2 metre radius, rotate at the same speed and let the hammer go the 20% increase in length translates to a 20% increase in release speed   of 12 m/S  so the result is then 720  ((10*144/2 ) or a 44% increase

3.       My third attempt I make a   1.5 metre radius, still achieve the same rotational speed so the hammer is now released at 15m/S  answer = 1125  or a 225% increase for a 50% increase in linear speed.

The example above clearly demonstrates that the relationship is not proportional and that is what the table shows, The lighter mass further out due to the use of the M5 pressure plate has a higher angular velocity at the same rotational speed and therefore has more energy than the heavier combination with the mass at a smaller radius.  

 

Sorry but you have a basic misunderstanding of what angular velocity is, radius has no effect on it, none, nada. That why it is measured in angle unit per time, not a distance unit.

https://en.wikipedia.org/wiki/Angular_velocity

Constant rotational frequency does indeed translate in to constant angular velocity. This is basic stuff, I'm quite surprised you've got so far through the theory to a result without understanding the basics.

That's why when you put in 5000rpm to one of the many, many online calcs, you get an angular velocity output (523rad/s) with no mention of radius anywhere.

https://www.calculator.org/properties/angular_velocity.html

The example demonstrates how you are getting confused. If you use linear velocity instead of angular, the linear equations work, however, earlier on you're mixing the two and using both rotational inertia, and linear velocity, this doesn't work. Which is what leads you to a conclusion where energy is not conserved across the examples (this should have been a red flag to you early on).

[Herbmiester 923]

So after all that the calculations were, whats the word, oh yeah wrong. 

[tawa 150]

17 hours ago, 3pedals said:

You are correct I have been mis-using the term angular velocity - I'm not confused , I  just shouldn't rely on my memory for  terminology and I should check to avoid misleading others. 

The linear velocity is what I used in my calculations  for the last column (as demonstrated in the example) I've also found the error(s)  in the spread sheet  

 The values are proportional  corrected table below - thanks 

Key point still stands - its not the mass that counts it is where it is located 

 

No worries, it's refreshing to see someone on a forum take on board a correction so full credit for that.

[wrs 120]

I'd be very interested since I have the flywheel... As before, my combo is the D1 flywheel with standard M3 Sachs clutch and pressure plate. It would be interesting to see how it stacks up. It's not something I put any research into before rushing out to buy, however, seeing this thread suggests I should have paid more attention to making a potentially better choice.

My engine is coming out on Monday next week to do all timing chains and rails/guides, leaking sump gasket, front and rear crank seals and new oil pump with locknut. If the D1 flywheel isn't all that good then I have the chance to correct it. If it is ok I'll leave it as-is. I have been considering going back to dual-mass to get rid of the chatter but I'm sure if I did I'd miss the fast rev pickup in 1st & 2nd.

[wrs 120]

Thanks, lower inertia than I'd expected. Explains why it chatters like a ....

Not only does it chatter at idle it also makes quite a strong meshing sound under low-medium loads under 3000rpm and sounds like it has rocks in it when going slow in 1st or 2nd gear. It chatters when cold or hot at idle but is much worse when hot. It's possibly worse on the S6S420G box too - doesn't it have bigger gears and bigger clearances than the 5 speed which makes it a lot more notchy?

However, it revs like crazy which I quite like.