Valve Run Out Gauge

Hi,

I am doing some cylinder head work, part of which will involve cutting / grinding new valve seats.

Having watched a couple of videos on valve seat cutting, the guy was checking the valve seats with a runout gauge. Can't find any in the UK but this is what they look like

I have a DTI gauge and thought maybe I could make something similar that would do the job? Need to make a pilot to sit in the valve guide, with a sleeve that spins around the pilot and with a DTI holder on it.

Has anyone on here ever made such a thing?

Thanks in advance

Such gauges were, and probably still are regul;arly used in industry.

Basically it is a pilot which is a tight fit in the valve guide, and bears on the end of it for axial location, which carries a DTI bearing on the valve seat.

Rotating the device will show the high and low spots (Eccenticity ) in the valve seat. Sometimes a basic handle is provided to facilitate rotation.

Rather than being available commercially, such things are usually bespoke items, made in the Toolroom of the factory concerned.

So, it looks as if you have just found another job on the Round Tuit. (Design and make a valve seat concentricity gauge. Two if Inlet and Exhaust valve guides are at different levels, or have different bores )

Howard

Looks like it.
I guess I could use an old valve.

Hang on, let's think about this. Perhaps the dial gauge is an example of PSONKing - an acronym used by an old mentor. The Pseudo-Scientification of No Knowledge turns up everywhere, when 'accurate' data is expensively amassed for little purpose - but it sounds impressive and 'scientific'.

What would you do with the quantitative data that a dial-gauge run-out gizmo would produce? All you can do is to take some, probably uncalibrated, corrective action to restore concentricity - perhaps by leaning on the seat cutter a bit. Then re-check, cut a bit more with more or different bias, etc. But the quantitative information is pretty redundant.

And the guides will have worn, so are you 'measuring' seat or guide errors? Answer - both...

You surely can get just as far by using the traditional engineer's blue and a valve with a freshly-trued-up seat. I like simple...

If the guide is worn, the datum point will not be consistent, so the results are likely to vary every

Taking a ridiculously extreme example, using a 6 mm pilot in a 1/4" guide will never give consistent and repeatable readings!.

These devices were used to check freshly machined, and unworn components.

In high volume production, this is a means of checking that components have not been mislocated by swarf trapped under the component, or that the tooling or machine has not deteriorated.

Howard.

The seats aren’t ground until the new valve guides are installed so the runout is used to check that a new seat has been evenly ground.

Yes, understood - standard practice to attack new seats only after new guides have been fitted (assuming guides are to be replaced). However, my comments stand.

The gizmo - or something equivalent - comes into its own in industrial QC if the part or assembly is being checked against allowable dimensions, but you don't need quantitative information. The traditional engineer's blue method gets the job done perfectly adequately.

That gauge is not going to show you what bearing blue will. The seat will be ground off the guides anyway. Being a 1 thou resolution , is very course for concentricity for valve guides to valve seats. I don't understand how you can't see if a seat is not evenly ground, sorry.

I'm struggling to see the point of measuring the run-out. If seat and/or guide have been replaced then the seat will need to be re-cut to ensure concentricity. I've repaired a number of cyclinder heads and never needed to make this measurement. The seat cutter will show and correct any error.

John

A lot of this depends on the accuracy of how you cut the valve seat, if you grind it how accurately do you dress the stone? how often do you dress the stone? what is the method of establishing the pilot to the valve guide? If you don't grind it but single point cut the seat again how do you establish the pilot.

First I would suggest you go and look at the Neway expanding pilot system for accurate guide piloting.

Secondly I would decide how you are going to machine the seats, grind, single point or multi cut. For me the only way is either a dedicated motorised single point bench system and as a second best the Neway system. Forget radiused formed seats, they are a waste of time and effort as has been proven by airflow testing, the only 'trick' system to improve airflow over a triple or quad angle seat is the 'rivlet' form behind the valve head over the last 10mm of the throat. If the method is correct the seat will be concentric, its simple! Grinding seats is the wrong process.

Martin

Often, in industry, to avoid having to lap valve to seat, there is a slight, about 1/2 degree, difference between valve and seat angles, to give a line contact for an initial seal.

Although this means that there is absolutely minimal area to conduct away heat while the valve is on the seat, the two hammer into greater contact, to provide the area for heat transfer and a longer leakage path.

At 4,000 rpm the valve spends very short periods of time on the seat anyway. About 1/4 of a millisecond each rev of a four cylinder four stroke!. So probably a lot of the heat is conducted away up the stem and into the guide. via any oil film. Maybe, the oil from the valve gear helps to cool the end of the stem.

Makes you realise that valves have a pretty tough life, and yet, thanks to the materials, survive for an INCREDIBLY long time..

Howard

Howard,

As 4000 rpm equates to 15 millisecs per rev and in a 4 stroke engine a cycle therefore takes 30 millisecs (for the 2 revolutions) the valve will spend a little more than a ¼ of a millisecond on its seat!

Given a typical road tune engine valve period is 250 degrees this will equate to 250/720 * 30 milliseconds opening period (about 10.42 milliseconds) hence leaving 30 - 10.42 = 19.58 millisecond on its seat. Even a race engine with 330 degree valve period will have 16.75 millisecond 'seat time' at this engine speed.

The Neway valve seat cutters are offered in 91 degree included angles, giving as you say, a ½ degree top locator to the valve / seat interface. This is then widened to about a 1mm (0.040" ) seat width during the lapping process.

Single point cutters with powered drive are available from Hunger GmbH - these are also available with 45½ degree blades for ensuring 'top location' seats, 30 degree included angle and 150 degree blades for triple angling.

We employed an 87% to 92% throat to seat constriction with 12 'rivlets' of approximately 1mm deep x 2 to 2.5mm wide. This was developed using low valve lift air flow measurements taken around the valve periphery and produced around 20% increase in low lift conditions with much increased swirl function. Another air flow trick was to produce the valve with a 'disruptor' on the back face of the inlet and on the front face of the exhaust. These again increased low lift flow figures as the disruptor worked in a similar fashion to the Gurney strip used on aerofoil wings to reduce drag.

Now work out how a late 1990's F1 engine managed to have 28mm valve lift and was able to run at 20,000 rpm, valve period was quite mild though at around 270 degree period!

Martin

I mmade a bvad approximation that a valve would be open for about 180m degrees.

The longer the open period, the less time spent on the seat to rid the heat collected, which makes life very difficult for the Exhaust. The Inlet, on a naturally aspirated engine does pass ambient air (If it has not acquired too much heat during passage into the cylinder through the induction tract.

On a turbocharged and intercooled engine, the charge is probably at 150'C by the time that it reaches the valve!

But this is relatively cool compared to exhaust temperatures of 650'C and upwards under full load rated speed conditions.. And these are for commercial engines, not tuned for racing or rallying, where specific outputs can be doubled, although at the expense of life probably.

(Ferrari's ideal racing would, he said, disintegrate as it passed the chequered flag! ).

And all valves expose a large area, to collect the head, to heat during combustion, so cooling is importance.

And who knows what shape the valve seat takes under F L R S conditions!

Howard

We ran aluminium inlet valves and titanium exhaust valves in the F1 engine (helps with that 28mm lift! ) but valve temperatures were reduced by the applied coatings (not ceramics by the way, which did not work, the coating needs to be intermolecular ). Intake temperatures were at sub zero in a 25C ambient when we ran carbon fibre intakes, this caused ice formation and insipient misfire due to fuel vapourisation issues, so we had to revert to aluminium tracts to get some heat into the fuel air mixture to aid the 'vapour gas exchange' process. Exhaust gases run at approx 950C in the exhaust port just after the valve head, turbos can get to over 1080C for short periods. The piston is the hardest working component in the engine though, through the cycle it can go from over 450C to less than 150C in 3 millisecs, the limiting factor is the average temperature through the cycle has to be below 320C for RR56 or about 360C for MPL 3 alloy.

Good point about valve seat distortion Howard, we often saw power figures deteriorate under prolonged full load runs due to seat distortion. A short timed fuel pulse was added to 'cool' the engine internals. This was a big problem on the Mulsanne straight at Le Mans in my days working on WSC cars, far more than at any other track, cured when they decided to put the roundabouts into the straight!

Martin

Yeahbut…. The sodium in the valve stems also transferred heat from the head to the stemms for dissipation via the guides. Not as simple as a morris minor, for sure.