Taming the bark of an Avenger

[TorqueMaster]

Folks,

I am looking for interested volunteers to print and test the two-part detachable LDC shown below.  The STL files can be downloaded from this link for 7 days: 
https://we.tl/t-osYckrIoxV

The design uses interrupted threads to attach the external LDC to the in-shroud 2-stage air stripper section. 

To fit the female thread inside the shroud, I had to scale that aspect down to 85% the size of the one Bob printed two days ago.  So, now the threads are a nominal 16 mm OD with a 2.55 mm pitch.   

The most significant aspect is that the front and rear sections of the thread are now integral to functional in-shroud and external LDC sections.

The thread pair still has a boatload of clearance to avoid the need for "shaving off a few tenths" before the parts will mate.

EDIT:  I check STL files for integrity to see if they will open with Cura.  The image below indicates the orientation on the print platen that is required.    I don't know if there is a reason to avoid printing them side by side.  Perhaps they are stronger or straighter if printed one at a time.

Looks good!

My first print will just be the ends, to make sure they fit ok.  So far you've been spot-on, but just in case...

I find myself wondering whats better to inside the shroud -- strippers or other elements?  Not being judgy, I truly don't know.  There isn't a ton of space so it should be put to best use.  Or nothing at all if the purpose is to test the effectiveness of the external part... 

Another, perhaps silly idea, is to intentionally discharge some air through a whistle of some sort.  I'm thinking very, very high pitched, but who knows -- if it sounds like a whistling bottle rocket I may have just invented the hula hoop!  "Never underestimate the stupidity of the American consumer."

It's cold and raining.  I'm printing not testing.

[subscriber]

crank it up like a joe six-pack or dial it back like an enthusiast would?

For consistency of testing, we should be the same...

We are all enthusiasts here.  I am not going to dictate settings, as long as the ones you use are consistent and clearly stated.  I would like your Avenger to reach around 30 FPE as a general benchmark.  It should be quieter doing that at a reg setting of 2000 PSI, compared to 1000 PSI, because the valve has to close later when shooting at low pressure.

Normally, getting good power and air efficiency are the goal.  Sometimes max power is the goal.   Shooting at max power is going to make more noise.  If you can suppress that, more power to you, but that is probably not the goal of the person who started this thread.

[TorqueMaster]

crank it up like a joe six-pack or dial it back like an enthusiast would?

For consistency of testing, we should be the same...

We are all enthusiasts here.  I am not going to dictate settings, as long as the ones you use are consistent and clearly stated.  I would like your Avenger to reach around 30 FPE as a general benchmark.  It should be quieter doing that at a reg setting of 2000 PSI, compared to 1000 PSI, because the valve has to close later when shooting at low pressure.

Normally, getting good power and air efficiency are the goal.  Sometimes max power is the goal.   Shooting at max power is going to make more noise.  If you can suppress that, more power to you, but that is probably not the goal of the person who started this thread.

Ok, I'll keep 1750 but throttle it back to the +0 to +3 range.  I should still be near the previous 29.45 FPE.

Ok new subject -- either I'm on crack or these parts have left hand thread!  The lug part for sure.  My .28 layer height makes seeing the threads harder on the 1mm pitch threads, but I think they are too...

Also it is under-rotating more than the last one.  I estimate it is rotating about 75-80% as far as it should, compared to 90-95% on the first one.

[subscriber]

Thanks for printing the parts and evaluating them, Bob

As usual, you ask good questions.  While the idea of testing only the relative effectiveness of an external LDC has merit,  the right way to do that is to cut the shroud flush with the barrel, and to attach the LDC directly to the barrel.

My goal is not to design the "best external LDC".  My goal is to design a good system that works in conjunction with the stock parts.  The Avenger lacks a primary air stripper to fully utilize the rear shroud space.  I started with correcting that in a manner that I believe should improve grouping ability. 

An empty narrow diameter shroud without air stripper will tend to have the air leaving the barrel muzzle pass the pellet on the way towards the end of the shroud.  Air passing the pellet in close contact is undesirable, because it tends to make the pellet wobble.  So, if one is trying to make a good system, then the air inside the shroud should be managed.   Especially that near the barrel muzzle, where the air has significant energy.

If you want an empty shroud version of the design to test, don't speculate; describe what you have in mind and I will provide the print file.

Getting back to the interrupted thread design; I originally intended to place that outside the shroud so that I could keep the diameter of the parts up for strength and stiffness.    Then, I remembered Matt saying that he thought the coupling should live inside the shroud, so I started with that.  I also chose an external LDC design that does not use a lot of plastic.  If that fails to keep alignment then using the smaller diameter thread won't stand a chance with larger diameter or longer external LDC designs.

I still have the larger interrupted thread design.  I just have to imbed that in the appropriate part designs.  The only downside is that the in-shroud part then has to grow in length by the coupling length.  Not a deal breaker, if it proves necessary, because the smaller diameter one is not up to the task.

[subscriber]

Bob,

Close to 30 FPE is perfect for this exercise.

I have not changed the 1 mm pitch threads in a long time.  If those seem left-handed, you may be having an incorrect experience

The larger threads started with a right-hand spiral.  Not sure if that got flipped, but your impression of the M24 x 1 thread might be insightful.

My impression of the larger test print you did was that it over rotated by 3.6 degrees.  So, I made a change to correct that, and make it under rotate by a similar amount.  The idea being that it would loosen up with use.  So, starting with some under rotation, rather than over rotation.  Also, when the coupling is attached to larger parts, you ability to tighten it would seem to favor starting a little short, rather than long.

EDIT:   You may be right about the interrupted threads acting left handed; despite how I intended to construct them.  Will have to see if there is an easy way to correct that, after your test pieces prove your impression right.  In that case, what I thought was over rotation may have been under rotation.  At least, the latter is an easy adjustment.

[TorqueMaster]

Whoops.  In message 235 I said

"Happy to help...until I find a new rabbit hole to dive into.  They are as hand tight as I can get them given the geometry.  It is slight undertravel.  So slight that it is fine now, and may be even better if there's more to grip, and as the parts wear.  And printed at a finer layer height may move it either way."

So it was always undertravel, and now it's even more undertravel.  Your correction did work...

I'm going to look at the STL file on the 24x1mm threads.  Running my thumbnail around the print and looking at the leading thread are both telling me it's Lefthand.

EDIT -- confirmed -- both the lug threads and m24 threads are lefthand.    This is on the newest 2-piece pair you posted.  I haven't looked at any others. 

[subscriber]

Thanks Bob,

If both large and small threads are left handed, then I somehow mirrored the part.  Which is what I will use to correct it

What about the front section?  I think its thread helix is the same as the rear one...

[TorqueMaster]

I've just been looking at the outer threads of both parts.  The inner threads on the interior piece-- it is hard to tell.  I can't see them straight-on.  The outer threads on the exterior part seem to "smoothly" rotate counter-clockwise into the inner piece, so I'd have to guess everything is left-hand. 

[subscriber]

Bob,

I have the mirrored parts.  Stand by...

[subscriber]

Thanks Bob for catching that mix-up.   Sorry for wasting your time.

The timing of the interrupted threads is now nominal in the 3D model - see images below.

Link to download the correctly mirrored thread STL part files:  https://we.tl/t-YCcVMzLDf1

[TorqueMaster]

No problem, glad it's sorted out.  I've been using these prints to help fine tune my PETG printing.  Instead of straight 0.28mm I'm trying a variable layer height that goes thinner on complex areas, and up to 0.28 max.  Also tweakting temperatures and speeds.  Seems to be working.

I looked at the parts files I printed previously, and they had reversed 24mm threads too.  I was apparently "cross threaded" -- perhaps that term applies more than ever in this situation -- and surely that contributed to my blow outs.  I'll skip the electrical tape and bet I'll be fine now...

[TorqueMaster]

Reprinted the connecting parts of the newly posted RH thread pieces.  All threaded sections printed at 0.12mm layer height on both pieces.  On first assembly, the inner part was over-rotating about 50%.  But then, with more disassembly/assembly, it got looser, and made it completely past that lug, and is stopping on the next one, barely overlapping it 10-15%.  I'm wondering if my changing from 0.28 layers to 0.12 caused enough extra/less gap to change the alignment that much?  This 0.12mm layer print should be closer to the design dimensions...  Also thinking making it a full thread rather than interrupted would eliminate the issue.

The 24mm threads are still lose for me.  Closer inspection shows if I use Cura's "inclusive" setting, it doesn't increase the threads OD, just widens some of them.  So I am using the normal "middle" setting.  I did some measuring, and have found if I print the part at 102% on the x and y, the threads are a great - dare I say perfect-  fit -- even better than the stock cap, which spins on very easily and has lots of wiggle room. 

I determined the part file had a thread OD of 23.45mm.  My 100% print came out 23.24.  PETG is expected to shrink about half a percent (to 23.33) so I'm a hair smaller than expected.  The stock cap thread OD measures between 23.50 and 23.62.  When I print X and Y at 102% I get a part that measures 23.70 to 23.75, and fits great.  One internet source says the 24x1.0 thread OD should be between 23.794 - 23.974, so I'm still shy of that but fitting fine, probably the slop of plastic printing helping out. 

I know you reduced the threads for others, and it seems I need them enlarged...so I am not sure where the answer lies.  If I print XY at 102% and Z at 100%, I do not change either thread pitch, just the diameters.  The thick inner thread will gain a little extra clearance, so it should still be ok.  So my intent is to print the inner piece with X and Y at 102% from now on, and you can leave the part dimensions unchanged.  LMK if that makes sense to you.

[subscriber]

Thanks for the feedback, Bob

It just goes to show that there is not one recipe that works for all printers, materials or print profiles.

If you have a working solution, then perhaps I don't need to change anything.  Else, I can make adjustments.    I never felt good about removing half the material from the M24 x 1 thread.  An extra 0.1 mm total clearance fine, but not half the thread area...

If you have a working solution, is the interrupted thread good enough to shoot with?  Is it aligned?  Does it stay tight?

Else, this seems like people requesting "improvements" they did not really want.  Otherwise, with the effort that you and I have put into it, this might have future usefulness in some related application.


EDIT:  At some point diameter adjustments in the print profile may prevent the the rear part entering the shroud, or slipping over the barrel.  It that is the case, then each dimension may need tweaking individually.

[TorqueMaster]

It's fun to figure out how to make this work, glad to help.

I did read back and saw how much you took off to match the thingiverse design.  About 0.4mm total OD, or slightly more, if I read it correctly.  Which if added back...would print almost perfectly for me at 100%.  Actually, it would widen the threads too, which might cause me the same problems as the others.  Anyway, my 102% solution is fine by me.

So, I do not really have a working model.  I have only printed the threaded portions for testing.  The interrupted thread is only overlapping the lug about 5-10% now, so I have not printed the entire parts.  I'm lobbying for full threads rather than interrupted, so regardless of printer settings, and how well it all lines up when rotated, there will be full engagement. 

I did some comparing of my android apps to the apple noish app, and with one I'm within 2-3db, so I'll keep on with the Android.

I should still be able to test the ones I printed, just have to secure them with tape.

[TorqueMaster]

Saw your edit -- my 102% still fits in the shroud.  The hole the barrel goes into will be slightly larger than before too, so no added interference there either. 

[subscriber]

Bob,

I created a version of the interrupted thread LDC and in-shroud air stripper that have full threads. 

I also added material to the M24 x 1 threads to make them stronger.  They are still smaller than nominal, but 0.4 larger in OD than the last version.  The last image shows the difference in thread profile.  The blue area is added material, over the Thingyverse thread profile.

Try printing at 100 % dimensions and see what happens.

The STL files are here for 7 days:  https://we.tl/t-W7G9KCjGYY

[TorqueMaster]

Printing at 100%.  I just looked at the STL, and my tools show the new thread OD is 23.70mm, just 0.25 bigger than the 23.45 saw before before.  Is my STL measuring tool wrong?  23.70 should print about 23.50 to 23.60, which is about the same as the loosely fitting stock cap.  It will be interesting to see if the widening of the threads snugs it up. 

On the previously tested part, I measured the 102% print before insertion at 23.70 - 23.75.  Just now thought to measure AFTER insertion, to see if it got deformed.  Not much really.  It's now about 23.70 all the way around.

[TorqueMaster]

As before, I printed off just the threaded sections to test fit.  For unknown reasons, my printer prefers printing the tall skinny 24mm threads more than the thickened ones.  Still works, just looks ugly -- some missing bits of thread here and there and little blobs scattered around.  This has been my experience with PETG -- rarely a "perfect" print. Perhaps higher end PETG behaves better than the low end hobby stuff I am using.

The 24mm threaded part goes easily and steadily into the shroud, no wobble, seats firmly and snugly.  No complaints.  The coarse threaded outer piece spins into it super smoothly, seats well, and tightly, and there's even a satisfying little "pop" on the last few degrees of twist, sort of locking it there.  I think it is the Z-axis seam/seams of the parts passing each other at that point, purely a lucky occurrence.  The visible gap between the two parts is perfectly uniform to my eye, all the way around.  The co-axiality of the two pieces should be very good.

I am printing the complete parts.  I'll post a followup at some point. I may get some outdoor testing done tomorrow.

Pic of the latest threaded parts test:

[subscriber]

Thanks Bob,

The last M24 thread OD measures 23.69 mm in my CAD.  I actually am not driving the OD.  Rather, I started with the nominal M24 thread, and then shaved off a number of .05 mm increments from it.

A nominal M24 thread has a smaller OD than 24 mm on the male thread; and a larger one than 24 mm at the root of the female thread.  Just how much the OD deviates from the nominal depends on the size, pitch and class of the thread.  It is always in a direction to provide clearance, to avoid interference.

[TorqueMaster]

The full parts printed fairly well.  I think Cura (I'm using an older version) does not work quite correctly with the variable layer heights -- it appears to cause more layer gaps than when sticking to a constant layer height, some of which coincide with layer height changes.  So these were printed at constant layer heights.  Inner part fits shroud very well.  Outer part fits inner part very well.  Feels like neither will rattle loose over time, but only way to be sure is keep a set installed and not uninstall it.

The pair seems to be very well aligned.  I rolled the assembled pair on a countertop and could not see any out of alignment wobble.

Sound testing -- not much good news here.  I tried the iPad and an Android, at various distances (from 6 inches to 10 feet) and settings and do not see what I would call significant sound level changes.  I definitely hear the changes.
The best results seemed to be the iPad NIOSH "peak" which I believe uses "C" weighting.   And the Android app's "max" -- no documentation, estimated to be updating the screen around 10Hz, so ballpark 100ms window.  Measured more or less inline with and 10 feet downrange from the muzzle.  Average of 3 to 5 shots  .22 Avenger around 1900psi, zero turns on spring, approx 935fps with 14.3gr CPHP, approx 26.0FPE.

Configuration tested --  iPad Peak -- Android max -- how it sounds
No cap                     -- 106.9       -- 71.4            -- very loud
Stock Cap                -- 105.9       -- 70.3            -- loud
Air stripper only        -- 103.6       -- 71.2            -- loud, noticeable "crack"
Stripper + Buckrail    -- 102.3       -- 68.8            -- major improvement, can hear hammer spring twang
Striper + MaximFatso-- 103.0       -- 66.6           -- same as buckrail, maybe a tiny bit quieter
Stripper + Plume       -- 103.1       -- 66.2           -- same as fatso, "seems" like the best of the 3

Without hearing all 3 in quick succession, it is hard to pick a best one, but I'll go Plume.  It's staying on the gun for now.  The numbers do not change as much as I'd expect for the sound levels I'm hearing.  Maybe "peak" is correct, but not meaningful.  "Max" may be diluting a 10ms (or whatever) burst over 100ms (or whatever) time frame...

Other notes -- the Fatso seemed to be negatively affecting POI.  Several shots went way off course, as much as 1 foot, at 25 yards.  I was concerned for the welfare of my electronics!  I didn't hear any pellet contact in it, don't see any evidence of it.  The others seemed to be putting pellets right where I was aiming them.