Forum » General Tuning Discussion » Forced Induction Plenum and Valve Spring Sizing

Forced Induction Plenum and Valve Spring Sizing

General Tuning Discussion

Forum Posts

Courses

Blog

Tech Articles

Discuss all things tuning in this section. News, products, problems and results. 

= Resolved threads

Page 1
Author
201 Views

I have a twin turbo 3.0l v6 Ford Duratec powering my Noble M400. The engine is fully built with forged internals, the heads are ported with Inconel high temp valves, stiffer valve springs and higher lift cams etc. The plenum is stock Ford with a Bosch motorsport dbw throttle body. The valve springs that are fitted to the heads are 70lbs seat pressure at a fitted length of 36mm.

When the engine is running at 1.6 Bar boost at around 6500 rpm the power drops off and seems to hit a wall. Even when more boost is added it makes no difference. At lower boost levels the engine revs comfortably past 7500 rpm .

There are a few cars in the UK running this spec engine and turbos. They all behave in the same way at higher boost and at 6500 rpm, this is around the 730hp level. One theory was that the plenum was limiting the flow to the engine. One friend had a new billet plenum designed and manufactured by a motorsport design company here in the UK, initially the new plenum did not make much of a difference and performed very similar to the original Ford item, then the the design company increased the volume of the plenum by increasing the size of end cans. Once this change was made, the engine revved comfortably past 7500rpm at higher boost and made more power to over 900hp.

We always thought that the plenum was the limiting factor, however recently talking to a tuner he suggested that increasing the size of the plenum volume could be masking another problem. He said that they had the same issue on the modern Audi RS3 / TTRS engine and initially changed the plenum but it ended up being an issue with valve springs and the stock Audi plenum was good for over 1200hp.

He suggested the following:

1. If the size of the plenum runner per cylinder is greater than the size of the inlet valves combined then the plenum should not be the restriction.

2. The issues that we are seeing is most likely related to pulses or disturbance in the air flow being caused by valve float. Then by increasing the plenum volume this is damping the pulses and hiding an underlying issue.

When I look at the simple calculation for the inlet springs based on 1.6 Bar boost they seem fine based on :

35mm inlet valve = 1.37 inch diameter, 1.49 sq Inch surface area

Pressure at 1.6 Bar boost = 23.2 PSI

Spring pressure minimum to overcome boost = 1.49 * 23.2 = 34.57 PSI

I also note that the area allowed on the back of the valve is probably larger than reality due to not taking in to account the reduction in surface area of the valve stem. So taking to account the quick calc the seat pressure of 70lbs is stronger than the pressure or 34.6PSI being applied on the back of the valve.

So based on the above, is there a method of modelling the correct valve spring seat pressure required in a forced induction environment to take in to account float and also the scenario with potential pulses in the airflow caused by float?

Any help would be appreciated

Many thanks Ed

So, if the springs were sized for a normally aspirated application, then it needs the spring force just to follow the cam and not have valve float. If your boost reduces the effective spring rate, then you would expect valve float to occur at higher RPM. So, I think you need to increase the valve spring rate by at least the mount of your maximum boost.

Now another solution might be to apply a pre-load on the springs. So you install shims to reduce the installed valve spring height (be careful that the springs can't hit coil bind at maximum lift). I would expect that pro engine builders do experiments like that to determine what the correct spring required, or what the installed height needs to be to avoid valve float.

The engine in this application has always run as forced induction and the springs run fine at lower boost levels. The springs are uprated from the normally aspirated donor engine in the original crate Ford engine. The challenge is when I run more boost we get a scenario that looks like valve float or maxing out the plenum.

As I have 70lbs springs fitted I'd like to understand what size do I need based on the math / modelling and if my current springs are marginal at the current levels of boost. The reason I am looking for the theory behind this is that stiffer springs are not readily available for the engine so it is not so easy to try a different spring rate, whatever I need I will need to have manufactured.

I did look at fitting a shim as a quick fix to increase the spring rate but unfortunately the lift of my cam does not allow for this and is very close to coil bound at full compression of the spring.

It does look like Supertech makes a 83 lb version as well as the 72 lb version (if I've gotten the engine right):

https://www.supertechperformance.com/p197300-sprk-fe20be-2-24-conical-valve-spring-kit

Seems like a conversation with the experts would be in order.

Good Luck!

Thank you David, yes I did see those. Unfortunately mine is the earlier 3.0 Duratec, the heads are different on the 3.5 litre with longer valves due to the VVT and no hydraulic lifters. The fitted length is different to what I need by 1.2mm that will increase the seat pressure further (which may actually be a good thing).

I was hoping to find an expert on here for the calcs and then to look for a spring to fit the spec.

Hi Eddie.

You might be facing valce spring surge issue. I do not believe that plenum size would be causing such a big restriction - even if it would that should still react to the boost increasing unless there is another problem with exhaust back pressure. Hitting 6500 RPM limit on power seems more like a mechanical issue - surging springs may be a good one to check on first, they can be easily resonating.

Plenum size should be not less than combined diameter of valves multipled by camshaft timing factor as not all the valves are open at the same time so that effects the total orifice used at any one time.

As for stiffer springs formula you already made proper calculation - it is area times pressure vs spring installed height pressure. The only one thing to remember is that apart from the force counteracting on top of the valve against spring we also need to consider the force created by valve acceleration ( valve mass times valve acceleration equals force)that would make it try to bounce back off the valve seat. So you would want to reduce the valve mass to reduce that force or install stiffer springs.

Hi Shota,

Thank you, I see that you concur with the suggestion is that the plenum size is most likely not the issue and in principle my calculations are in the right ball park for the sizing of the seat pressure.

In terms of dealing with valve spring surge I will be unable to change the valve mass greatly as the valves are already Inconel for heat resistance. The interesting point is what impacts surge on the valve springs and whether the air volume around the spring impacts or dampens the oscillations. For example when my friend changed the volume of the plenum the problem went away. Unfortunately the cost of the plenum was significant.

As I ideally want to stay with the stock plenum, the next question is how do I address the valve spring surge. Is the only way to replace the springs with a different rate or construction?

Would it be trial and error or is there a rule of thumb?

The current springs rate is 242 lb/ft, is this reasonable for a turbo v6 application and what would be too high based on experience?

Well, the lowest spring rate for turbo applications I have seen is 267 lb/ft whilst they are normally around 280-300ish... Looks like yours are a bit too soft. As Davis already said you can try to use shims to insert underneath of spring to increase the force on valves but you will need to consider valve maximum lift figure to avoid coil bind.

Attached Files
  • Screenshot_20210303_105533_com.android.chrome.jpg
  • Attachments may only be downloaded by paid Gold members. Read more about becoming a Gold member here.

OK, just to go over some of the fundamentals, as they're sometimes overlooked.

With valve springs there are numerous things to consider, in no particular order...

Spring rate - the force required to compress it a specific distance.

Spring diameter - obvious, but it may belimited by other factors, such as interference with other engine parts.

Spring free length - how long the pring is when fully extended under no load.

Spring compressed length - how long the spring is when fully compressed and the coils in full contact, but we're more concerned with

Spring minimum installed length - this is the minimum length the spring can be compressed to while maintaining a 'safe' clearance between the coils - usually this will depend on the spring dimensions and some sources suggest 60 thou' as a minimum, but some have run smaller springs as close as 30 thou', 0.75mm. This is important as coil bind will cause instant failure.

Spring resonant, or harmonic, frequency - this is often overlooked but can be critical as it's possible for the spring to physically 'bounce' off it's seat. If there are signs of fretting of the seat, this is often the cause as it allows the spring to move on it's seat. This is also one of the reasons for using 'beehive' springs, besides clearance, as the frequency varies along their length and so dampens harmonic development.

We then have...

Seat pressure (actually a force, but convention...) - this is the force the spring is applying to the seat-retainer and is how much force is holding the valve closed. This is usually the problem when increasing manifold pressure.

Fully open spring force (AKA "over the nose") - this is the spring force at full lift. It is an important factor in avoiding valve 'float', where the valve train is no longer controlled, and which can cause canshaft/follower damage and valve bounce, which is where the valve physically hits the seat so hard it bounces off it and, when you consider the piston is chasing the closing exhaust valve, can lead to piston to valve contact. It can also cause fatigue failure and the valve head physically breaking off the valve stem which is not a good thing.

In this application, we have two things to consider - the rpm range and the manifold pressure - and how they will affect the springs chosen to meet all the critera for those changes and still meet the physical design limits. First the current spring parameters have to be established, including installed height, then what is going to be required for their replacement.

It has been a while, but IIRC, total nose force is going to be proportional to the rpm, as will the increase in seat 'pressure' be proportional to the increase in manifold pressure.

With all that information, you should be able to figure out what you need and so be able to check it against the valve springs listed by the various camshaft manufacturers. You 'may' be able to simply shim the springs, but if you get it wrong...

That said, another option, if you have the financial resources, is to reduce the valve train mass by using lighter follower buckets, spring retainers/caps, thinnner or hollow stem valves, or even swapping to titanium valves.

Hi Eddie - a bit of a stab in the dark, but I think plenum size could be ruled out almost completely because the volume of air ingested is not changing with boost, only the mass. If the plenum volume works at your base boost for your target RPM, it should work fine at higher boost because the VE of the engine has not changed, it is gulping the same volume but of denser air. This means it must be a valve control issue. Experts can correct me if I have cocked this up.

Cheers, Andrew

As other have said, fluid dynamics change due to charge density is relatively modest, ideal runner will be slightly different dimensions but not enough to matter.

The plenum volume masking the issue is because it is removing the intensity of intake resonance which is obviously coinciding with valve/spring/lifters harmonic at that engine speed. As already said, lighter valvetrain components, dual valvesprings/different installed high/rate etc will be required to control it properly. Most of us don't have tools to effectively analyse the valvetrain dynamics to any significant extent.

Thank you for your responses, the logic behind the questions from my original post is making more sense now.

In terms on the plenum the volume of the air passing through remains the same as the VE has not changed in the engine, it is limited by the rpm and the size of the bore / stroke and the valves. What changes is the density of air when compressed. So the plenum is not causing an issue but changing the volume of the plenum seems to be removing the intensity of the intake resonance, this could be moving the resonance issues to another frequency that may or may not impact on performance.

So the issue looks to be related to the intake resonance coinciding with the valve train harmonics at that rpm as mentioned above. The seat pressure of the valve springs is fine as long as it overcomes the pressures being exerted on the back of the valve plus those created from the acceleration of the valve train through the combustion cycle.

So the next step is to look at how can I change the harmonics of the valve train. As the valves are already Inconel and upgraded, the simplest / lowest cost option seems to be to move away from a single spring and to go to a dual spring and also to increase the spring rate to about 300 lb/ft from the 242 lb/ft fitted today.

Thanks again

Eddie

It would be good Idea to review all reciprocating parts and replce them with ligher ones where possible... The picture attached clearly explaines why.

Attached Files
  • Screenshot_20210306_184837_com.android.chrome.jpg
  • Attachments may only be downloaded by paid Gold members. Read more about becoming a Gold member here.

Ah, thanks Shota, I was about to comment on the spring rate going up as the square of the increase in RPM. This would mean that the fully open valve would have the spring force increase by ~33% going from 6k5 to 7k5, assuming all else being equal, with the seat pressure (force) change for the increased boost being added to that. You may need to go to double valve springs, and these are commonly a light interference fit and contra-wound to reduce harmonics.

Ah, re-reading this, there doesn't seem to be much mention, if any, about the exhaust side. The pressure built up in the exhaust manifolds can be several times the intake side pressure, especially if a smaller than optimum turbine and/or housing is used. You may be correct in your assumption that the pressure build up is the problem - just the wrong set of valves.

Some of you may find this footage of interest - I would particularly draw attention to the spring physically leaving the seat in the second, and just how much the valve stem is moving around in the third.

https://www.youtube.com/watch?v=XfFQMbKWhKg

https://www.youtube.com/watch?v=50F-haEs3pg

https://www.youtube.com/watch?v=_REQ1PUM0rY

Im running the same engine and am friends with the Author and we have discussed this at length before but never came up with a definitive answer (we are not experts, just owners trying to make the engine as efficient as possible just for fun)... there is quite a bit of additional background information that could be added based on my experiences but here are the most significant.

1. I rev the same engine to 8100rpm on high boost and it made over 700bhp for 1500rpm right to redline.. but hits the boost wall as mentioned above.. i push the rpm harder than others with the same power.

2. the stock plenum is originally from a 4x4 auto truck that redlines at 6100rpm.. long runners, tiny plenum, horrible bends in the runners etc.

3. actuator pressure sees the torque drop at 6500rpm, where is goes past the inlet goes past its optimum, i used boost to make power high up.

4. When I stripped my engine the exhaust valve guides were all very badly worn, one seat was actually loose and all the cams had impact marks one of which was 1mm deep (cams were scrap)

5. Inlet valves had no issues at all

The heads are rockers with hydraulic lifters and whilst I had closely monitored oil pressure in my Motec i had come to the conclusion the head had oil issues at high rpm and have moved to a cam over valve with solid bucket head with a dry sump.... I also developed a carbon plenum with runners calculated for 7100rpm peak and ITB's... Ive not run the engine yet as i'm still developing the heads.

So yes it could be the exhaust valves that re floating (causing damage wear my case and restricting boost for everyone).... HOWEVER the other inlet that was developed unleashed 200bhp with not a significant increase in boost or any other changes.

What we don't understand is if valve float on the exhaust is restricting boost and thus, how can changing the inlet solve that?

Is it simply that we have solved a separate issue that the inlet is actually restrictive and thus we see a power gain... but if we started pushing the boost up again we would encounter the wall again?

There seems to be a bit of confusion there, probably on both our parts, but my thoughts.

The 'boost" as seen in the intake plenum has little direct affect on the exhaust pressure in the manifolds - this is primarily down to the resistance to the gas flow by the turbine, spinning the impeller, and housing - this can be several times the pressure in the intake. This pressure is working against the spring to hold the valve open.

The second thing is the actual valve spring, which needs to be strong enough, at the various points of the valve travel, to hold the valve train together. The higher the rpm, the more spring force required and, from what damage you report, I assume it was on the back side of the lobe?, it would seem you weren't using enough spring force (pressure) and the follower was crashing into the camshaft - the followers are also scrap, but since you're going to solid it doesn't matter, anyway.

I would expect an improvement in intake and camshaft efficiency to actually drop your boost, as there will be less resistance to filling the cylinders, but because the exhaust pressure is also a function of the volumes of gas, it may be expected to increase - but a larger housing will help reduce that and aid higher rpm power. However, there will be numerous factors, all interacting, and it's difficult to say where 'the wall' will be, if anywhere.

You are correct, my cams had damage after lobe and previously i had attributed this to oiling issues in the head at the high rpm i was going to as at 8000rpm i was on the limit of the oil pump and it runs hydraulic lifters... however it might have been due to valve spring, I have a much larger turbo housing that anyone else (as i run a very larger single turbo and they run small twins) but i'm also reving higher, so....

I guess it is possible what happened is the inlet is a restriction and by removing that the boost was lowered for the same power (EGT also dropped significantly), so as the boost was targeted the same as before a significant power increase was seen?

We never measured EBP previously so whilst i will know what it is in the future we wont know what it was previously on mine or others packages.

I don't think I will get a definitive answer as I have changed the head design fundamentally and will be using double springs and have also upgraded the inlet so if i see gains I wont know which was the original cause of the restriction.

Ah, that pretty much confirms it was, at least in part, a valve spring problem. What happens is the valve springs can't control the inertia of the valve train and the valve/follower 'floats as it loses contact with the lobe, only to smash back into the back of the lobe as the spring manages to push it back into contact. In some cases this can be relatively mild, but in other cases it can not only damage the camshaft(s) and followers, but it can result in them bouncing several times and even continuing through the exhaust closing point, with the valve physically bouncing on the seat.

It also means the exhaust valves aren't able to seal the cylinder against the higher pressure exhaust gases which will reduce the intake charge of air-fuel.

You were actually very lucky as this will often result in piston-valve contact, bent valves, and even failure of the head - and a destroyed engine, as the piston is chasing the closing exhaust valves, and if they're lagging...

I knew the engine was sick so towards the end i didn't use it hard.... Whilst it didn't blow up, nothing was salvagable from my engine except ironically the valves which I cant use in my new heads.

I started again from a new crate engine and all new forged internals :(

My new cams machined from blanks actually arrive tomorrow, i had to spec these from scratch as the heads are fundamentally different to the stock ones (solid buckets rather than rocker and hydraulic lifter).