Is solid roller lifter loading really the highest at idle?

General engine tech -- Drag Racing to Circle Track

Moderator: Team

540 RAT
Expert
Expert
Posts: 733
Joined: Thu Feb 08, 2007 7:21 pm
Location: Southern California

Is solid roller lifter loading really the highest at idle?

Post by 540 RAT »

BBC solid roller lifter failure discussions on the Forums, usually get a lot of interest. So, here is yet another related topic we can chew on. From time to time during those Forum discussions, it is stated that "The load on solid roller lifters is the highest at idle". When this is mentioned, it is usually said matter-of-factly, as if its common knowledge. However, some folks may have missed the memo spelling that out. And since that thinking is obviously the total opposite of high rpm, then by extension, that would mean that the loads would be at their "absolute highest" with the engine simply being rolled over with a wrench. But that might seem counter intuitive to some. So, let's take a look to see if that statement is really true.

When one guy who said that, was asked "why" he said that, his response was that an old cam guy had told him that back when he was young. He didn't have an explanation as to how that could be the case, he just took what he was told at face value and was adamant that it was fact. But without any support data to back up the statement, he was no help in determining if it was true or not.

When another guy was asked "why" he said that, he stated it was because, "The lifter is lofted "off" the nose of the lobe at high rpm, thus leaving no load on the lifter. But that at idle, the lifter would always be under load from the spring. So the loads are highest at idle". He would be correct about the lifter having no load for that brief instant in time that it was in the air free of any contact, "if" you were floating the valves. But he didn't bother to take into account the whole picture. Because of course any relief the lifter feels when floating in the air, is way more than made up for by the hammer blow shock load it would feel when it comes crashing back down. And the end result is that his scenario is much much worse for the lifter at high rpm, not better. And he never even considered properly setup engines that don't float the valves. So, his response was also of no help in determining if the statement is true or not.

Then some might figure that they could support the claim from what we see in actual usage, since the solid roller lifters in lower rpm street driven motors, seem to fail more often than those lifters in high rpm race motors, in spite of most lifters now being pressure fed. But if you consider absolute time/mileage differences between these engines, then even "weekend only" driven street Hotrods get way way more total time/mileage on them than most any race car. And from the failed roller lifters that I've inspected, the number of pressure cycles on them reached the metal's fatigue failure point, which supports the idea of extra usage taking its toll. In addition, early failures in street motors may also be partly because many street guys favor the thicker 20W50 oil in their motors, which can contribute to roller skidding on the lobe. And roller skidding reduces the metal's fatigue life, which can bring on failure sooner than it might otherwise occur. Many racers like to use very thin oil for max hp, but that can also help keep the rollers spinning without skidding, thus helping to extend the metal's fatigue life. So, the added mileage and thicker oil likely play rolls in the overall picture of accounting for street roller lifters apparently being more prone to failure than race car roller lifters, without having anything to do with the rpms involved. Obviously, with all the variables that come into play, the whole issue is rather complex. And then of course there are some street car roller lifters that do indeed seem to go for a long time, which would also contradict that original claim about the lower rpms having the highest loads and being worse on the lifters. Also, if you look at the bell curve generated when evaluating the failures of any part, you'll see that for otherwise identical parts, most have reasonable lives, while some fail early for no apparent reason. Though it is probably due to minor manufacturing differences. So, any attempt to try to use anecdotal evidence to prove that the loads are highest at lower rpms, would really only be guessing, which leaves you with nothing concrete to support that original claim. That being the case, real world observation still doesn't help us get to the bottom of the "rpm vs loading" question either.

So, my take on all this, is to take a more scientific look at things to see what's actually going. If you look at just an instant snapshot in time, of the force from the spring on the lifter, while on the nose of the lobe, the load is the same for, rolling the engine over on the engine stand, at idle or at max rpm (assuming a proper setup with no valve float). In this hypothetical situation you are only looking at the force on the lifter due to spring pressure. But of course in a running engine it is a dynamic situation. And then you would be accelerating the whole mass of the entire valve train, PLUS compressing the valve spring. Big difference. Here's a little more detail on all that. One of the most fundamental equations in Physics is F=MA, which stands for force equals mass times its acceleration. Now if we apply that equation to one of our roller lifters as it accelerates up its bore, along with the whole valve train involved, we see that we can omit the mass for general discussion since the mass of the lifter and its valve train of course does not change. So, we are left with force or load (same thing) on the lifter, being proportional to the acceleration of the mass of the whole valve train. Obviously as the rpm climbs, the lifter accelerates up its bore, along with the rest of the valve train, quicker and quicker. And the equation tells us that the only way to accelerate all those parts quicker and quicker is for the force or load on the lifter to increase in order to do that. Since we can't argue with the laws of Physics, the load on the lifter is clearly the LOWEST AT IDLE (or really just rolling the engine over with a wrench) and the HIGHEST AT MAX RPM.

If you just do a simple sanity check and think about it, it makes perfect sense. You are really whacking those valves open extremely quickly at high rpm, and moving the mass of the entire valve train in order to do that. But at idle or simply rolling the engine over, things are quite gentle in comparison. An analogy might be a bullet being fired out of a gun. If you want to accelerate the bullet quicker down the barrel to get more velocity as it leaves the end of the barrel, you need more gun powder behind it to provide more force. Same concept for accelerating the mass of the lifter and valve train faster, it simply takes more force or load pushing on the lifter. So, the windup is that the load on the lifter is clearly the highest at max rpm, NOT at idle. This shows that the original claim of the load being highest at idle, is just not true.
blykins
Guru
Guru
Posts: 2128
Joined: Mon Aug 27, 2007 9:59 pm
Location: Louisville, KY

Post by blykins »

I agree with your conclusion that the load on a solid roller lifter is not highest at idle.

I see a lot of people asking the question, "Well hydraulic roller camshafts last for 100's of thousands of miles in certain cases. Why is a solid roller any different?"

The answer is lash. A hydraulic roller doesn't have valve lash. The lash in a solid roller setup is really what causes the damage. The lifter is not always in perfect contact with a load on each side of it, so it essentially gets continually hammered.

I think the amount of stress is directly related to rpm and I think it goes a lot deeper than just F=ma. I agree that the load on the lifter is increased by its own acceleration, but I think that's minimal compared to the fact that the lobe basically falls out from under the lifter. When that happens, the lifter is unloaded on one side, is acted upon just by the force on one end (the pushrod end), and then is acted upon by a great shock force when the lifter comes back into contact with the cam lobe.

In the end, I agree with your final result. However, it's been so long since I finished my Mechanical Engineering degree that the detailed laws of inertia, momentum, and fatigue stresses will escape me until I really get back into it again.
n2omike
Guru
Guru
Posts: 1067
Joined: Sat Sep 02, 2006 10:56 pm
Location: West Virginia

Post by n2omike »

They just get the least oil at idle, as they rely on what is flung off the crank. SLOW idles can kill cams/lifters.
PackardV8
Guru
Guru
Posts: 7619
Joined: Sun Jul 30, 2006 2:03 pm
Location: Spokane, WA

Post by PackardV8 »

One cruise night can put four to six hours on the valve train. Depends on the length of the line at your strip and how many runs you make, but say ten minutes max per run, that's the engine hours equivalent of 24 to 36 runs. That's a whole season of racing in the north country.

What kills roller lifters on the street is running a race profile and springs. Can't have race performance and street durability. We can pick the fly poop out of the pepper about oiling, clearances and the other small details and we'll move the failure mode a bit in or out a few hours. Bottom line, there are no known lifters which will live indefinitely in street use when subjected to race spring pressures and race intensity.

GM has been selling a 7,000 RPM pushrod hydraulic roller engine with 100k mile life expectancy. Would appreciate one of our cam guys comparing that lift, rocker ratio, duration and intensity with what is considered the max solid street roller specs for even 20k mile durability.

thnx, jack vines
Jack Vines
Studebaker-Packard V8 Limited
Obsolete Engineering
Barbapapa
Expert
Expert
Posts: 995
Joined: Sun Nov 12, 2006 9:04 pm
Location: Dacula, GA
Contact:

Post by Barbapapa »

PackardV8 wrote:
GM has been selling a 7,000 RPM pushrod hydraulic roller engine with 100k mile life expectancy. Would appreciate one of our cam guys comparing that lift, rocker ratio, duration and intensity with what is considered the max solid street roller specs for even 20k mile durability.

thnx, jack vines
Well it does have titanium intake valves which not only have less mass for the lobe to move but require less spring to slap closed. I like the idea of Ti valves with lighter springs and have seen no drawbacks when using them in a street car.
Mike S.
blykins
Guru
Guru
Posts: 2128
Joined: Mon Aug 27, 2007 9:59 pm
Location: Louisville, KY

Post by blykins »

What do you consider race springs? >500lbs open?
540 RAT
Expert
Expert
Posts: 733
Joined: Thu Feb 08, 2007 7:21 pm
Location: Southern California

Post by 540 RAT »

n2omike wrote:They just get the least oil at idle, as they rely on what is flung off the crank. SLOW idles can kill cams/lifters.
That's not true with modern lifters, most if not all, have pressure fed oiling. For example, when priming my 540 with the intake off, oil floods out of the lifters and engulfs the cam lobes. And that's only from my drill motor. Of the failed lifters that I took apart, not one of them showed any signs at all of oil starvation. There was no galling/ scraping or heat induced discoloration. They all failed from pitting/flaking of the metal surfaces, which is a textbook example of metal fatigue failure. Loading is a primary factor in determining a materials fatigue life. Oiling was simply not involved in the failures.
TRN
Member
Member
Posts: 163
Joined: Thu Nov 09, 2006 3:30 pm
Location: Quincy, WA

Lifter load

Post by TRN »

The load on a lifter is roughly the sum of the spring force and F=ma force. If you look at the acceleration plot for your lobe, you can see where acceleration is positive and where it's negative.

This is where the rumor that load is highest at idle came from. Typical profiles have high positive accelerations at low lift, where spring force is lower, and the acceleration is usually negative over the nose, where spring force is high. Thus the over the nose force IS highest at idle. The problem with this logic is that roller lifters fail due to peak loads, not total loads. The highest load is encountered where the sum of the spring force and the F=ma force is highest, which will be somewhere on the opening ramp.

Your comment that the over the nose force will be the same at any RPM, is flat wrong. I suppose you may have a cam with zero acceleration over the nose, for a brief instant, but this is not typical in my experience. I should qualify that remark by stating that my understanding of the state of the art in camshaft design is about 30 years old.

It is possible to simulate all this, and as an aside, this is what motivated this farm boy to go to college. First came calculus and physics, then statics, dynamics, and diff eq. I bought a PDP 11/05 to run the calculations, and ended up with a degree in electrical engineering after using that computer! My mother said I was wasting my life playing with cars!

The tools available today are vastly superior. You should probably re-evaluate your old timers that don't know where this idea came from, most of my understanding came from great mentors, and the selection and care of your mentors will change your life.
540 RAT
Expert
Expert
Posts: 733
Joined: Thu Feb 08, 2007 7:21 pm
Location: Southern California

Re: Lifter load

Post by 540 RAT »

TRN wrote:The load on a lifter is roughly the sum of the spring force and F=ma force. If you look at the acceleration plot for your lobe, you can see where acceleration is positive and where it's negative.

This is where the rumor that load is highest at idle came from. Typical profiles have high positive accelerations at low lift, where spring force is lower, and the acceleration is usually negative over the nose, where spring force is high. Thus the over the nose force IS highest at idle. The problem with this logic is that roller lifters fail due to peak loads, not total loads. The highest load is encountered where the sum of the spring force and the F=ma force is highest, which will be somewhere on the opening ramp.

Your comment that the over the nose force will be the same at any RPM, is flat wrong. I suppose you may have a cam with zero acceleration over the nose, for a brief instant, but this is not typical in my experience. I should qualify that remark by stating that my understanding of the state of the art in camshaft design is about 30 years old.

It is possible to simulate all this, and as an aside, this is what motivated this farm boy to go to college. First came calculus and physics, then statics, dynamics, and diff eq. I bought a PDP 11/05 to run the calculations, and ended up with a degree in electrical engineering after using that computer! My mother said I was wasting my life playing with cars!

The tools available today are vastly superior. You should probably re-evaluate your old timers that don't know where this idea came from, most of my understanding came from great mentors, and the selection and care of your mentors will change your life.
OK maybe you didn't really read my words carefully, so here they are again:

If you look at just an instant snapshot in time, of the force from the spring on the lifter, while on the nose of the lobe, the load is the same for, rolling the engine over on the engine stand, at idle or at max rpm (assuming a proper setup with no valve float). In this hypothetical situation you are only looking at the force on the lifter due to spring pressure.

Here I'm only referring to the load due to the spring, while acceleration or lack thereof isn't even included in my statement.

And you said:
Your comment that the over the nose force will be the same at any RPM, is flat wrong.

So, feel free to knock yourself out trying to disprove what I said. Remember force is not acceleration, either positive or negative. They are two entirely different things. Force is in pounds, while acceleration is in feet per second squared. But then you "should" already know this.

Then I went on to say:

But of course in a running engine it is a dynamic situation. And then you would be accelerating the whole mass of the entire valve train, PLUS compressing the valve spring. Big difference.

And here, I am not disagreeing with your statement that said:
The highest load is encountered where the sum of the spring force and the F=ma force is highest, which will be somewhere on the opening ramp.

I think you said that very well, and I totally agree with that. But I don't see where anything I said runs counter to that, since I didn't specify the exact point where sum of the combined forces is the greatest. I only referred to them combining.

You also said:
I suppose you may have a cam with zero acceleration over the nose, for a brief instant, but this is not typical in my experience.

Ummm, every lifter, let me make that clear, EVERY lifter will have zero acceleration on the nose at max lift. It will come to a dead stop, just like a piston does, and then it changes direction. It has to stop (zero acceleration), in order to change directions. You say this is not typical in your experience. Please clue us in as to what you have experienced that operates differently.

And you said:
This is where the rumor that load is highest at idle came from. Typical profiles have high positive accelerations at low lift, where spring force is lower, and the acceleration is usually negative over the nose, where spring force is high. Thus the over the nose force IS highest at idle.

Here you first say that the highest load at idle is a rumor, then you say the highest force IS at idle. I'm not following you, you contradicted yourself. So, what are you trying to say here, you can't have BOTH ways, pick one.

I don't THINK we are really that much in disagreement. Maybe we just don't exactly understand what the other really is trying to say.
xenginebuilder
Expert
Expert
Posts: 529
Joined: Sat Aug 18, 2007 8:57 pm
Location: MN, USA

Re: Lifter load

Post by xenginebuilder »

540 RAT wrote:Ummm, every lifter, let me make that clear, EVERY lifter will have zero acceleration on the nose at max lift. It will come to a dead stop, just like a piston does, and then it changes direction. It has to stop (zero acceleration), in order to change directions.
Don't misinterpret lack of valve motion or relative motion at max lift for lack of acceleration forces. With rare exception in automotive cams, the design of the profile acceleration stays negative across the nose. The velocity is going from positive to negative the entire time, and the 0 point is just one point on that slope. As RPM increases, the forces on the cam and lifter increase at the opening and closing side, and decrease between those points as the spring absorbs more and more of the momentum, until the spring cannot keep the valvetrain in consistent contact. Maybe UDHarold or CamKing can weigh in here and explain it better, but the acceleration at max lift is definitely not zero. BTW, the acceleration forces on a rod and piston at TDC is what pulls these parts apart at speed and fatigues them.

I guess an analogy would be to imagine rolling backwards in your car at 25 mph and suddenly dropping it in forward gear and mashing the gas. The rear tires would spin and you would decelerate to zero speed and then accelerate in the forward direction (assuming you did not spin out and end up in a ditch LOL). The acceleration you feel in your seat is relatively constant even though your direction and speed changes the entire time. Your passenger, who has his eyes closed, would not be able to tell the difference
If you lend someone $20.00 and never see that person again, it was probably worth it.
User avatar
Wolfplace
Guru
Guru
Posts: 3580
Joined: Fri Sep 24, 2004 10:39 pm
Location: Mendocino County, Northern CA
Contact:

Re: Lifter load

Post by Wolfplace »

xenginebuilder wrote: Don't misinterpret lack of valve motion or relative motion at max lift for lack of acceleration forces. With rare exception in automotive cams, the design of the profile acceleration stays negative across the nose. The velocity is going from positive to negative the entire time, and the 0 point is just one point on that slope. As RPM increases, the forces on the cam and lifter increase at the opening and closing side, and decrease between those points as the spring absorbs more and more of the momentum, until the spring cannot keep the valvetrain in consistent contact. Maybe UDHarold or CamKing can weigh in here and explain it better, but the acceleration at max lift is definitely not zero. BTW, the acceleration forces on a rod and piston at TDC is what pulls these parts apart at speed and fatigues them.

I guess an analogy would be to imagine rolling backwards in your car at 25 mph and suddenly dropping it in forward gear and mashing the gas. The rear tires would spin and you would decelerate to zero speed and then accelerate in the forward direction (assuming you did not spin out and end up in a ditch LOL). The acceleration you feel in your seat is relatively constant even though your direction and speed changes the entire time. Your passenger, who has his eyes closed, would not be able to tell the difference
=
First a small preface,,,,
While I hesitate to get involved with this thread because the last time I had an opinion on the subject of lifters which was contradictory I was accused by Mr Rat of making things "personal" & being a "troublemaker" & "unpleasant",,,,, :lol:
Secondly, I am not a cam designer nor am I an engineer so take my opinions for what they are
The way I understand this crap :wink:

While I think I understand what you are saying,,
To me the part in bold makes no sense & is contradictory unless you are referring to seating velocity on the closing side??

Assuming you were still referring to what happens over the nose,,
How can the negative forces go higher except in a negative direction
In other words the load goes down

While I understand the increase in force or load in a positive direction
Why would the load go up in a slowing positive or negative direction?

The acceleration across the nose is not negative, it slows, goes to zero & then goes negative as you stated prior to the bold part & obviously you cannot get from positive to negative without passing through zero at some point.
And if done at to high a rate obviously bad things happen,,,

And the load or force on the follower will most certainly decrease & the higher the RPM or the higher the negative acceleration the more the force or load will be effected
It will in fact decrease to zero if you go to far
If this were not so you could run checking springs on your engine & do away with them 1200lb suckers,,, assuming you did not have other issues on closing.

If we use F=MA
And you make either quantity zero what is the force?
If you make "A" a negative quantity what happens to the force?
Mike
Lewis Racing Engines
4axis CNC block machining


A few of the cars I have driven & owned
A tour of my shop
The Dyno
And a few pics of the gang

"Life is tough. Life is even tougher if you're stupid"
John Wayne
xenginebuilder
Expert
Expert
Posts: 529
Joined: Sat Aug 18, 2007 8:57 pm
Location: MN, USA

Re: Lifter load

Post by xenginebuilder »

Wolfplace wrote:While I think I understand what you are saying,,
To me the part in bold makes no sense & is contradictory unless you are referring to seating velocity on the closing side??
I think the problems we are all having, is defining the terms in relationship to the valve train motion. Velocity and acceleration are not interchangeable.
Velocity is speed, acceleration is rate of change of speed. From the maximum (positive) velocity on the opening side to the maximum closing velocity (negative, opposite direction) the velocity is constantly changing. Because the definied direction is from positive to negative, the rate of change, acceleration, is also negative, and not zero. Acceleration only goes to zero when there is no change in velocity. All the force you put into the valvetrain on the opening side, has to be opposed by an opposite force of equal magnitude. The opening (and closing) forces are almost unlimited, since the parts are in compression and only limited by material strength. All that opposes that is the valve spring.

One more example, and I'll call it a day. If you imagine a motocross biker riding over a hill. As he increases his speed of approach to the hill, the suspension compresses more and more as he hits the hill, and he unloads the suspension more and more over the top, until eventually, he spends most of the time in the air going over the top. As long as he lands on the other down slope side before he reaches the bottom, all is well, and he stays in control. He goes a little faster, and lands on the flat ground, he breaks his neck, while bouncing a few times. The rider is the lifter, the hill the cam, and gravity the spring. Everything is the same except the speed of the bike.
If you lend someone $20.00 and never see that person again, it was probably worth it.
desoto30

Post by desoto30 »

Not trying to be a smart a$$ here. Some clarification of terminology is in order. xenginebuilder,speed is speed. = distance over time. velocity is speed in a given direction. a small but important difference,so speed & velocity are similar but none the less different.Acceleration = change of speed or velocity.
User avatar
Wolfplace
Guru
Guru
Posts: 3580
Joined: Fri Sep 24, 2004 10:39 pm
Location: Mendocino County, Northern CA
Contact:

Re: Lifter load

Post by Wolfplace »

xenginebuilder wrote:
Wolfplace wrote:While I think I understand what you are saying,,
To me the part in bold makes no sense & is contradictory unless you are referring to seating velocity on the closing side??
I think the problems we are all having, is defining the terms in relationship to the valve train motion. Velocity and acceleration are not interchangeable.
Velocity is speed, acceleration is rate of change of speed. From the maximum (positive) velocity on the opening side to the maximum closing velocity (negative, opposite direction) the velocity is constantly changing. Because the definied direction is from positive to negative, the rate of change, acceleration, is also negative, and not zero. Acceleration only goes to zero when there is no change in velocity. All the force you put into the valvetrain on the opening side, has to be opposed by an opposite force of equal magnitude. The opening (and closing) forces are almost unlimited, since the parts are in compression and only limited by material strength. All that opposes that is the valve spring.

One more example, and I'll call it a day. If you imagine a motocross biker riding over a hill. As he increases his speed of approach to the hill, the suspension compresses more and more as he hits the hill, and he unloads the suspension more and more over the top, until eventually, he spends most of the time in the air going over the top. As long as he lands on the other down slope side before he reaches the bottom, all is well, and he stays in control. He goes a little faster, and lands on the flat ground, he breaks his neck, while bouncing a few times. The rider is the lifter, the hill the cam, and gravity the spring. Everything is the same except the speed of the bike.
And the rate of change is what needs to be controlled
The velocity was only brought up because this is what is important at seating & I was not sure if you were including this in your example
Or,, the acceleration needs to be lowered to improve velocity at seating,,,

Read the part I quoted in bold
What I am saying is if you are referring to the area around the nose of the lobe where the acceleration slows, stops & changes direction
(positive/zero/negative)
Then the forces or loads on the follower & lobe most certainly do change

Again, if they did not assuming this is the only issue discussed you could run the Valvetrain with checking springs to most any RPM :wink:

And I don't see the spring as absorbing the momentum, I see it as controlling the change in momentum
On this maybe we are saying the same thing differently??

Damn I am glad we have folks here who design this stuff for us,,,, :lol:
Mike
Lewis Racing Engines
4axis CNC block machining


A few of the cars I have driven & owned
A tour of my shop
The Dyno
And a few pics of the gang

"Life is tough. Life is even tougher if you're stupid"
John Wayne
ZIGGY
Guru
Guru
Posts: 1297
Joined: Mon Mar 05, 2007 3:15 am
Location: Dallas, Texas

Post by ZIGGY »

desoto - I don't know #@&* but I think you mean vector.
Post Reply