What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Vernatherm Woes

hgerhardt

Well Known Member
Now that PlaneKool is working so well, I’ve been venturing into warmer locales where pre-A/C, I just didn’t bother to fly. But now, oil temps are making regular excursions to 225*+ F. I set up a test by using my Bluetooth HVACR refrigerant line temp probes to measure oil cooler (OC) inlet/outlet temps and compared those to indicated on my G3X. During warmup, I saw that the OC inlet was measuring within 3*F of indicated, which gave me some confidence of the BT probe’s accuracy with the less-than-ideal clamp arrangement on a hex nut. Found that indicated temp was only 6*F cooler than OC inlet temp and that there was 16F delta of OC inlet/outlet, indicating the Vernatherm (VT) wasn’t doing a good job. Problem found! Or so I thought…

BT temp probes.jpg

Vernatherm Testing
Set up a water bath using a sous-vide heater and a height gage with a home-made ring gage, the dimensions of which I got from Lycoming SI 1316A and tested the 12-yr-old, 750-hr VT together with the original to the engine from 1980 and found the newer one didn’t even extend far enough to completely close the port. Bought a new VT, retested and made a bar graph of the results; points within the upper left quad are good. The red vertical line is design temp to close the port fully and the blue line is extension length to close the port. The original 1980 one was better than the 2008, but the new one is dramatically better. Problem solved! Or so I thought…

vernatherm testing.jpgVT chart.jpgSI 1316A pic.JPG

Vernatherm Seat Repair
Installed the new VT and flight-tested it. Not a huge amount of improvement (7F delta of OC inlet vs indicated, where previously it had been 6F). Thought about it and went back and looked at the 2008 VT and to me the seat was possibly leaking at the VT cone. I had checked the VT seat in the engine with a (low-quality) endoscope before I replaced the VT and it looked OK… however, after inspecting the 2008 VT, I saw the cone didn’t have even wear on its face. Time to clean up the seat per that same SI 1316A. The tool called out to remachine the seat is available to buy for… $2100! You gotta be kidding me… fortunately I had a couple of dead VT’s available, so I made my own from one of those plus $40 of parts from McMaster and 10 min on a lathe.
Cleaned up the seat and checked for any loose chips and went for a flight.

vernatherms.JPGseat cutter.JPGvernatherm seats.JPG

Results were better, but still not where I thought it should be, with 11F inlet/indicated delta. Time to do some more thinking…
Which brings us to: Ducts Matter
 
My compliments Heinrich. That extension vs temperature graph is very interesting. And I like your seat resurfacing tool too!
 
My compliments Heinrich. That extension vs temperature graph is very interesting. And I like your seat resurfacing tool too!

Thanks, Dan!

If anyone wants to borrow the seat cutter, it's yours to use for shipping both ways.
 
Excellent pictures and troubleshooting. This is especially interesting to me to see the wear on the cone as I had just recently examined mine and it is a nice comparison. On mine, you could barely see the markings on the cone but it did show complete mark all around.
If I had learned as much in school as I am learning here, I would be a top notch professor by now.

Many thanks
 
Heinrich, very good work, thanks for sharing.

A couple of questions, if you could be so kind.

1. It appears that the two older VT's you tested had a shorter initial measurement - is this because the ring you have dropped into the wear groove on the conical tip? Were their room temperature OALs likely the same?

2. Do you happen to have any OAL measurements at room temp?

I measured my 19 year old/1700 hour VT this year, but I measured delta OAL instead of the method you used (see post 13 in this thread). This is probably reasonable as the cone on mine showed no wear, just a slight witness of where it contacts the seat. The change in length of mine from room temp to 185F was .19", and at 200F was .24", which was in spec so I put it back in. I plan this winter to do something very similar to what you have done w/r to the oil cooler ducting.
 
Heinrich, very good work, thanks for sharing.

A couple of questions, if you could be so kind.

1. It appears that the two older VT's you tested had a shorter initial measurement - is this because the ring you have dropped into the wear groove on the conical tip? Were their room temperature OALs likely the same?

2. Do you happen to have any OAL measurements at room temp?

I measured my 19 year old/1700 hour VT this year, but I measured delta OAL instead of the method you used (see post 13 in this thread). This is probably reasonable as the cone on mine showed no wear, just a slight witness of where it contacts the seat. The change in length of mine from room temp to 185F was .19", and at 200F was .24", which was in spec so I put it back in. I plan this winter to do something very similar to what you have done w/r to the oil cooler ducting.

I had truncated my chart because didn't think anyone cared about the coolest temps... below is the whole chart starting at 80F. The oldest ones start almost identically in length even with the big gouge in the oldest. And I can't remember where in the groove the ring gage sat. The new one starts .020 longer at 80F.

I did not measure OAL. And, I got my required length to seat the cone from SI 1316A. Where did you get yours from?

Looking at your measured lengths, your VT appears less good than my original 1980 one, which didn't actually pass, looking at the chart. Attached the raw data too with the lengths subtracted from 80F so you don't have to squint at the chart. ;)

VT chart2.jpg VT raw.JPG
 
Last edited:
No surprise, but the Rosta factory drawing and SI 1316A seem to be in general agreement.
.
 

Attachments

  • ScreenHunter_736 Oct. 22 22.04.jpg
    ScreenHunter_736 Oct. 22 22.04.jpg
    15.7 KB · Views: 256
  • Scan_0012.jpg
    Scan_0012.jpg
    37.1 KB · Views: 258
I had truncated my chart because didn't think anyone cared about the coolest temps... below is the whole chart starting at 80F. The oldest ones start almost identically in length even with the big gouge in the oldest. And I can't remember where in the groove the ring gage sat. The new one starts .020 longer at 80F.

I did not measure OAL. And, I got my required length to seat the cone from SI 1316A. Where did you get yours from?

Looking at your measured lengths, your VT appears less good than my original 1980 one, which didn't actually pass, looking at the chart. Attached the raw data too with the lengths subtracted from 80F so you don't have to squint at the chart. ;)

View attachment 3695 View attachment 3694

You can try something on your sub-standard units. On my 320 rebuild, the VT didn't pass. I fully disassembled it and found a decent amount of carbon deposits inside. After fully cleaning, it passed. I believe the reduced clearance from the deposits was limiting travel. Maybe worth a few minutes to see if you can revive one.

Larry
 
Dan, at what temperature is the measurement on the drawing of the VT depicting?

The line drawing includes a caption stating "0.160" Min, Travel Ref". As I read the drawing, it means a cold vernatherm should measure....

3.280 - 0.160 = 3.120"

...from the underside of a nominal 0.0625 thick sealing washer. I note all of Heinrich's vernatherms measure a bit short of this dimension at 80F. However, ring gauge sizing and any cone wear would complicate the measurement.

In a separate spec list, we have:

"Valve travel to be 0.160" minimum between 150F D/C and 185F U/C."

and

"Valve must close against seat between 183F and 187F."

I have no idea what is meant by "D/C and U/C". And I apologize to all about second hand data. The Rosta drawing was provided by one of their engineers under an agreement not to publish the document itself.
.
 

Attachments

  • ScreenHunter_737 Oct. 23 08.03.jpg
    ScreenHunter_737 Oct. 23 08.03.jpg
    7.9 KB · Views: 209
What causes the severe wearing on the cone?

Thanks Dan.

BTW, the attached article has an explanation for the cone wear, which seemed very mysterious to me until I read it. (Dan - I believe you sent me this some years back).
 

Attachments

  • Vernatherm Article-AOPA-April2005.pdf
    97.4 KB · Views: 2,348
The line drawing includes a caption stating "0.160" Min, Travel Ref". As I read the drawing, it means a cold vernatherm should measure....
3.280 - 0.160 = 3.120"
...from the underside of a nominal 0.0625 thick sealing washer. I note all of Heinrich's vernatherms measure a bit short of this dimension at 80F. However, ring gauge sizing and any cone wear would complicate the measurement.
In a separate spec list, we have:
"Valve travel to be 0.160" minimum between 150F D/C and 185F U/C."
and
"Valve must close against seat between 183F and 187F."
I have no idea what is meant by "D/C and U/C". And I apologize to all about second hand data. The Rosta drawing was provided by one of their engineers under an agreement not to publish the document itself.
.

Dan et al,
I think it's time to step back and analyze our long-held assumption, which is that if the valve moves .160" from 150F to 185F, it's OK. I think that's a bad assumption and here's why:

There are only two pertinent dimensions on that VT drawing:

1. The seating length at 185F of 3.280", which is measured from a specified dia on the cone to the engine-side of the sealing washer.
2. The minimum travel of .160" from 150F to 185F. (note there is no maximum, nor a specified travel)

Note that there is NO callout for length at any temperature other than 183-187F (avg = 185F).

Because it is difficult to measure a VT with a ring gage at 185F, pretty much no one bothers... which is how/why most people just measure the .160" growth and call it success/failure based on that. That's wrong.

Remember that the only thing that matters with a VT is that it closes the port in the engine at 185F. That is the specification. And just by measuring a travel delta from an unknown/unspecified cold length is not accurate.

Note in my 1st post the pic of the height gage zeroed out on the sealing washer. That is how the OAL dimension is specified and is how I measured them. Dan notes that all my measurements seem short at 80F, but that is only because he assumes the .160" travel downward from the design length at 185F is a spec, which it is not.

Below I've attached my measurement table with some added columns of length deltas at 80F and 150F starting temps. You'll see that my 1980 VT passes the .160" test but fails at the design length at 185F. (it actually moves .175 from 150F to 185F).

SO, based on all that verbiage, what to do for the average guy without a height gage? Well, it would be fairly easy to make a GO/NO-GO gage from a fender washer with a .560" hole placed on top of a 1.5 ID tube, 3.530" length. You'd cook the VT at the 185F spec and then quickly put it on a flat table, hex base down, and then place the tube over the top of it and then the washer on top. If the washer rocks, the VT is long enough. Drawing below.

VT dwg.jpg VT notes.jpg VT raw.jpgScreenshot 2020-10-23 145903.jpg
 
Last edited:
You can try something on your sub-standard units. On my 320 rebuild, the VT didn't pass. I fully disassembled it and found a decent amount of carbon deposits inside. After fully cleaning, it passed. I believe the reduced clearance from the deposits was limiting travel. Maybe worth a few minutes to see if you can revive one.

Larry

I found no carbon, but I did find some lead residue, but not enough to cause anything to stick (just a thin film). I cleaned the internals before testing.
 
I think it's time to step back and analyze our long-held assumption, which is that if the valve moves .160" from 150F to 185F, it's OK.

Good point, and your measurements prove it.

SO, based on all that verbiage, what to do for the average guy without a height gage? Well, it would be fairly easy to make a GO/NO-GO gage from a fender washer with a .560" hole placed on top of a 1.5 ID tube, 3.530" length.

Works for you and me, but the fabrication might be too much for the average builder without a lathe and/or suitable reamer. Within the degree of precision required, wouldn't an ordinary caliper be sufficient for overall length at 185F, measuring to the wear line on the cone?
 

Attachments

  • Caliper.jpg
    Caliper.jpg
    9.9 KB · Views: 190
Within the degree of precision required, wouldn't an ordinary caliper be sufficient for overall length at 185F, measuring to the wear line on the cone?

Good point. Looking at my length table, even that old 1980 VT would pass the length test at 190F, but not at 185F.

And the new 2020 VT grows .095" between 180F and 190F, indicating you need to be fairly precise with controlling water temp if you want an accurate length test. Looks like a good excuse to get a sous-vide cooker if you don't already have one!

That relatively huge growth delta within 10F also shows a caliper ought to be plenty accurate to measure length to the wear mark. However, that method would be hard to use to measure a new VT without a wear mark to verify it isn't DOA out of the box.
 
Back
Top