ET: Engine (Twin) Camshaft design

passenger0_0

Well Known and Active Forum User
Non-VOC Member
As others have detailed, the radiused follower can only be used with a sympathetic cam form .

Why do you say this Phil?

I'm only asking this because I have previously used very aggressive cams with radius followers of both linear and lever types, some in (relatively for old engines) high speed applications (10,000 rpm).
 
D

Deleted member 3831

Guest
Phil
the polynomial method of cam design does not necessarily result in a cam profile that has any form of negative curvature, it is a means of designing a smooth progressive transition from one linear design point to the next.
Also variously referred to as curve fitting, or splining, the method is most often applied using 5th order polynomial equations.
Attached is a screen shot of a polynomial curve applied to the linear points of a cam nose, which I trust will adequately illustrate the form.
Regarding the Picador cam, I am surprised to learn that tripple valve spring packs were used for this application, it sounds like a belt and braces arrangement to avoid any possibility of stoppage as a result of valve train failure.
Does anyone know whether any of these tripple spring packs still exist ?
 

Attachments

  • 5th order polynomial nose example.jpg
    5th order polynomial nose example.jpg
    59.3 KB · Views: 33

timetraveller

Well Known and Active Forum User
VOC Member
I think that we need to be careful with the use of two terms, polydyne and polynomial. The latter, as shown by Grey One, is a mathematical way of fitting a smooth curve to a series of points. Incidentally when I designed my own cams I used a spline fitting routine in ACAD to interpolate between the points. The polydyne, as I understand it, refers to allowing for a correction to the cam profile which will compensate for the dynamic bending and compression of various parts which connect the cam surface to the valve movement. I had thought that I would try to be a smart alec when designing the cams and allow for these flexibilities. Two things stopped this. The first was that when I tried to measure a deflection by levering the end of the rocker when the cam was on lift I could not detect any movement. Secondly, when I did the sums it turns out that at 6,000 rpm the maximum acceleration is equal to 200 g. That is one pound weight of parts acts as though it was 200 pounds weight. But at lower revs, where road engines spend most of their time, say 3,000 rpm then this force is reduced by at least a factor of four.
This leads on to the reason why valve springs have to be so strong. Bits of the valve train acting as though they had a mass of 200 lbs means using strong springs and there is a lot of misinformation about concerning valve springs. I have measured the spring rate of various types of valve spring used on Vincent engines. The typical dual valve springs have a rate (working from memory) of 100 lbs/inch. The specials, which were provided by Terry Prince to go with some of his cams, had a rate of 150 lbs/inch. I read several comments about these wearing cams and followers out. Finally, many years ago when many of us were trying to go faster, several people went for using BSA Gold Star valve springs. The ones I measured had a rate of 200 lbs/inch and did not have a reputation for premature wear of various components.
I think that the information that Picadors used triple valve springs is incorrect. These were tried by Vincents on their competition engines and found not to be necessary. Many years ago I knew someone who knew someone who claimed to have been involved with some of the testing of these at the factory. Sadly both are now dead and there is no way of checking what I was told. This concerned the testing of a single engine with the triple valve springs at 8,000 rpm for up to a day. When I mentioned this on the Forum some years ago this was dismissed at rubbish and I have now no way of checking whether I misunderstood or was misinformed about what went on. Certainly for road and racing use the Vincent factory gave up on the use of triple valve springs well before the production of Picador engines. There are some in the Club who have these engines unused in their possession and so it would be possible to check those engines. However, as with all things Vincent one cannot be sure that what one finds in one engine will be typical of what one would find in many. A single example is the use of Mk II cams in some Picador engines and the special cams in others.
 

vibrac

Well Known and Active Forum User
VOC Member
I have heard of the 8K for a day test on motorcycle engines at Vincents l from a guy who spent most of his time on the 'advanced brake' sponsored/supplied by the min of defence for the life boat engine quote 'not the old thing they used on M/C engines'
 

Phil Davies

Well Known and Active Forum User
Non-VOC Member
Passenger0_0 - "because I have previously used very aggressive cams with radius followers of both linear and lever types, some in (relatively for old engines) high speed applications (10,000 rpm)" -
Quite possibly, what I should have added to my original post was 'not unless the cam is specifically designed for a flat follower' - but I can see how my comment could easily be taken as scoping outside of Picador cams (which it was not intened to be), sorry about that.

Time traveller
I understood that back in the day (a long time ago as well as in valve spring tech terms), triple valve springs were used for damping the valve return motion, as back then, upping the spring rate was the first point of call for 'performing' engines, with possible consequent seat bounce otherwise.
I was under the impression that the due to the continious performance demand requirements picador engines used the lightning engine spec as a starting point, with changes (compared to a lightning) appearing as a consequence of the (pre MOD) testing for the running duration requirements, the cams and followers seemingly being an obvious one.
Going slighly away from my initial thread comment.
I too have used Gold Star springs in my race engine, I was always nervous of the increased valve train loading as a consequence, I had a cam follower snap (then jam against the cam and peel it open - oops!), but was unsure as to the real source of the failure (Stevenage age followers!).
I have to admit that I was always concerned about high spring rate on flat lever followers (snapped followers are not an uncommon story in non factory vincent racing history).
Nowadays all that is transformed by the availability of progressive rate springs that give unbelievably light seating pressures yet still control against wild cam forms.
 
D

Deleted member 3831

Guest
Interesting conversation.
As rightly pointed out , it isn't practical to use flat followers when using serious lift and duration cams as the point of contact 'jumps' from one end of the follower to the other so you loose your sliding contact and encounter wear issues.
The Indian cam system I mention also used a concave form, necessary to get symmetrical lift profile, and was on the acceleration limits as defined by AG Bell.
No matter how you look at it, when using pivoting cam followers, you always end up with the point of cam & follower contact moving all over the place so you end up with strange looking asymmetrical cam profiles to realise symmetrical valve actuation.

I must take you to task for remarks you are making re firstly, the flat follower and the point of contact which you say 'jumps' from one end of the flat follower to the other. This cannot happen unless the engine speed becomes so excessive that inertia of valve train components allows a break in contact. In normal use, the cam slides along the flat from start of lift to the end. The lift and duration are immaterial unless lift is such that the cam lobe overlaps the end of the follower flat
Secondly, your further remark that cam and pivoting follower contact 'move all over the place' is a similarly erroneous statement.
A cam in rotational contact with any pivoting follower will have a contact path which proceeds across the face of the follower, whether the follower be flat or a radius. It cannot move about, the cam rotation is continuous in one direction, and contact is maintained between cam and follower accordingly.
My above comments are applicable to all followers, flat or pivoted
 

passenger0_0

Well Known and Active Forum User
Non-VOC Member
Unfortunately this conversation Grey One is of no real value to most followers of this forum and this is my last positing on this topic. You are simply wrong with your comments as I will demonstrate below.

Using the flat follower mathematical model described in the textbook Cam Design and Manufacturing Handbook (2002, Robert Norton), shown by the first figure below.

1516330764187.png


Following their method of analysis you get reasonable results until the lift get high (for our geometry and base circle, over 0.400” lift). Following this analysis you get the following impossible cam profile shown below:

1516330798075.png


The irregular location of the contact point between the cam and follower occurs when lift gets too high, causing the problems I mentioned earlier. QED.
 

david bowen

Well Known and Active Forum User
Non-VOC Member
passenger o o, Knows what he is talking about,he has raced Vincent twins rides a Vincent on the road and if I need any fine engineering points,I can rely on him given the correct info
 

timetraveller

Well Known and Active Forum User
VOC Member
My twopen'orth here. I calculated the position of the contact point between cam and follower for a Vincent cam with a flat lever follower. That is the standard set up. It might be true to say that the statement '
'jumps' from one end of the flat follower to the other' is not correct but it is certainly true that the position changes direction rapidly and more than once during one cycle of the cam rotation. It does not proceed continuously from one end to the other and back again. Some of the positional changes are very rapid and that is why I do not believe that a roller based cam follower would be useful with the standard Vincent layout. The problem is that if a cam were designed to work with a roller follower, even one on a lever, the cam shape would be different and I do not have calculations for that.
 

greg brillus

Well Known and Active Forum User
VOC Member
The Horner brothers ran roller tipped lever followers on the Goodwood racer. I have seen pictures of them, but i don't have any to put on here.
 
Top