Logo
Builders' Plan Gallery  |  Hip Pocket Web Site  |  Contact Forum Admin  |  Contact Global Moderator
October 16, 2018, 12:20:58 AM *
Welcome, Guest. Please login or register.

Login with email, password and session length
 
Home Help Search Login Register
Pages: 1 2 [3] 4 5   Go Down
Print
Author Topic: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work  (Read 4478 times)
0 Members and 1 Guest are viewing this topic.
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #50 on: December 21, 2017, 09:48:50 AM »

I as an engineer myself I appreciate that Don Stackhouse takes the time to comment on my proposals. I actually love it when he say I'm all wet and then goes in to a light (no maths) explanation. As to Don having more important things to do than toy airplanes, we can say that about all of us.  This might have some therapeutic benefit as it allows for a distraction. I'm not going tell Don how to use his time.

Thinking that DJAerotech had to make compromises for production reasons, I too thought that I could get some weight out of the airframe with tapered parts. DOn came back and said what would work and why he didn't taper some of the structure. As for the carbon caps I might have tapered the doubler over the outer wing panels. But I'd leave the spar on the tip panels a constant cross section as the model would /will need to survive my ground loop landings. Shocked

As the wing joiner is not in between the spar caps, wrapped spars might not be advantageous. I too have concerns about the continuos load path across the dihedral joint. But DJAerotech does have you use a fair bit of glass across this joint.

Spring loaded surfaces. So the servo is constantly under load. This puts a constant load on the output transistors. I'm not aware of any servo amp being spec'd for a 100% duty cycle on the outputs.

The spar valley as I've drawn them are grossly exaggerated. The humps I sanded off were less that 0.2mm above the spar. The covering is to BE bonded to the spar cap. So the covering will make sealed compartments that will balloon when heated. If I need to release any covering I've have found the duct tape allows me the pull up on the So-Lite and free the trouble spot.

I'm using cellulose glue for its body as a filler and as it is easy to sand. I hope not to need to keep the iron dwelling on the surface of the rib as I shrink the covering.


Love the cling wrap idea to test the curves!

All the best,
Konrad


« Last Edit: December 21, 2017, 10:51:19 AM by Konrad » Logged

Cut it twice and it's still too short!
lincoln
Platinum Member
******

Kudos: 29
Offline Offline

United States United States

Posts: 1,752



Ignore
« Reply #51 on: December 21, 2017, 09:56:17 AM »

I didn't say more important than model airplanes, though I suppose you're right. What about the proposed three meter which should be available about the time fusion power plants come on line? (Or about the time I get my regular Chrysalis 2M done.) ;-)
Logged
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #52 on: December 21, 2017, 10:04:05 AM »

Thanks for the video link.  The model featured had a 'heavier-looking white' covering over the 'd-box' area.  What was it?
It was in the right place to help with torsional stiffness.
Don't know, I assume it is So-lite. I do know that this was the first time Don used So-lite and that he is still not comfortable using it. I've used So-lite a lot in my gummy band to R/C conversion and will say that it takes awhile to develop the nuances to use it. I don't like it as it isn't thermally stable and the adhesive (color) tends to delaminate in short order.
http://www.hippocketaeronautics.com/hpa_forum/index.php?topic=16387.0

BTW: I think the green and white one is Joe's. Don's I think is the blue and white one.
Logged

Cut it twice and it's still too short!
dosco
Gold Member
*****

Kudos: 6
Offline Offline

United States United States

Posts: 856



Ignore
« Reply #53 on: December 21, 2017, 10:05:12 AM »

Spring loaded surfaces. So the servo is constantly under load. This puts a constant load on the output transistors. I'm not aware of any servo amp being spec'd for a 100% duty cycle on the outputs.

Konrad:
There's been some conversation about this on RCG (could be on the Chrysalis thread, I forget, and you don't read there).

Consensus appears to be that in the neutral position there is no load that causes drain on the RX batteries.

With that said, I've never tried it (I'm guessing you have), so YMMV and all that.

Cheers-
Dave
Logged
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #54 on: December 21, 2017, 10:13:22 AM »

I didn't say more important than model airplanes, though I suppose you're right. What about the proposed three meter which should be available about the time fusion power plants come on line? (Or about the time I get my regular Chrysalis 2M done.) ;-)
I would love to see a 3 meter I'm bored looking at Bubble Dancers and AVA's all day long. Roll Eyes A 2 meter with articulated trailing edges would be fun.
But as the Bubble Dancers is such a power house I don't think there are the sales for a 3 meter kit. And as there are fine foam full house 2 meter plus gliders I don't think a sport full house woodie will sell well. So concentrating on the F3-RES format sounds like a good use of DJArotech's resources.

Fusion power plants come on line? Heck, I keep waiting for them to pop up in the labs.  They are always another 20 years out!
« Last Edit: December 21, 2017, 10:53:04 AM by Konrad » Logged

Cut it twice and it's still too short!
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #55 on: December 21, 2017, 10:25:49 AM »

Spring loaded surfaces. So the servo is constantly under load. This puts a constant load on the output transistors. I'm not aware of any servo amp being spec'd for a 100% duty cycle on the outputs.

Konrad:
There's been some conversation about this on RCG (could be on the Chrysalis thread, I forget, and you don't read there).

Consensus appears to be that in the neutral position there is no load that causes drain on the RX batteries.

With that said, I've never tried it (I'm guessing you have), so YMMV and all that.

Cheers-
Dave
Yes, I've tried them about 10 years ago, not that that means much. I fully admit I may have messed up on their implementation. I was using GWS and Cirrus  5 gram class servos. Pull pull has worked out much much better for me. I'm still open to the idea if the limitation I see can be addressed (constant load on the output drivers of the servo).  This load while it might not tax most batteries, in the past it has burned up far too many of my servos.
Logged

Cut it twice and it's still too short!
USch
Gold Member
*****

Kudos: 21
Offline Offline

Italy Italy

Posts: 952




Ignore
« Reply #56 on: December 21, 2017, 03:26:35 PM »

Hi Konrad, I love your approach with attention to every detail.
So to encourage the discussion a few points which hit my interest  Wink

In reply #29 you write about lighten the tip plate scavenging them. What I missed was an indication how much you saved. Did you weight before and after?
I try to weight every piece I put my hands on, mainly because I remember well a wing where I lost a few evenings punching holes in all ribs just to find out, at the very end, that it was not worthwhile because the net gain was only very marginal. Or on my actual build putting carbon cap strips only on the last centimetres instead from D-box to TE, difference on the whole wing 0,15g but loss of rib stability.

Reply #35, I try to not finish with carbon dust over balsa, maybe sanding to shape the carbon before mounting it on the structure. If not possible finishing first the balsa and giving a coat of sanding sealer BEFORE sanding the carbon. If anything fails never try to remove the dust with your fingers, you just push it into the wood grain. Either blow it away or take a brush and brush the dust off the balsa.

Reply #45, argh...on the argument of gussets we could discuss for days  Grin  rounded off gussets look extremely nice !!! But structurally/weight wise dont make sense. I made a sketch to explain my reasoning. The red part of the gusset is useless weight (wow.. at least some 0,01g). Structurally only the fibres which run completely from one end to the other are useful.

Reply #55, todays digital servos are nearly 100% of the working life under strain, constantly trying to maintain the position. If you remember the days of the introduction of digital servos and our surprise that these halved the useful battery time you understand what I mean.

Just some food for thoughts  Grin

Urs
Attached files Thumbnail(s):
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Logged

Fast up-Slow down
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #57 on: December 21, 2017, 06:02:35 PM »

Urs,
By my triple beam scale a little under 1.6grams combined weight savings. It looks like the added diagonals and gussets have added 2.2 grams. So, I'm in the red 0.5 to 0.6 grams. But as I've moved this mass from way out board and aft of the spar to inboard and ahead of the spar I'm structurally way ahead. In my flight test work when the structure guys started talking about the dynamics of flutter my eyes glazed over. I was shocked to learn that the 747 has 500 lbs of dead mass added to the wing tips ahead of the main spar, and that wing flutter is what defined the critical speed of the airframe.

On the other wing tip I masked the tip that I tried to protect from the carbon dust. I brushed off the carbon from the balsa. I then used silly putty to lift up any carbon in the wood pores.

The reason for the concave gusset is to prevent a stress riser. The sudden change in cross section made by the tips of the black triangle are known as force concentrators, not what we want to see in dynamic structures. Now I agree wood is not particularly susceptible to stress risers, but  most glues are. Also generally the harder or stiffer a material is the more likely it is to perform poorly in the presents of sudden cross section changes. The runout of the  red areas is very helpful (needed) in surviving the violent stress from flutter or anything that presents high cycle fatigue. I agree that statically or in low frequency environment the need to radius the corners might not be too important.

That's a good point about digital amp. But as the load is constantly from one direction it is the same transistor(s) that is on (sees the load). With a pull pull the load around center is cycled between both pairs of transistors. This lowers the duty cycle considerably (lowers the chance of runaway heating). Even with the switch frequency of the amps output I have found that the output of a digital servo overheats if a load is kept constantly on (high air-loads, stiff hinges or return springs). I'd think that to get an amp driver large enough to dissipate the output heat the weight would cost more than adding the other cable for a pull pull set up.  And with digital servo this would be even worse, with the way a digital amp works. I'd need to set up or see on a trace scope the power output of the amp with a spring return verses a pull pull set up. I think I know the answers, but a controlled experiment is always a good idea.

All the best,
Konrad
Logged

Cut it twice and it's still too short!
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #58 on: December 22, 2017, 09:39:53 AM »

Hi Konrad, I love your approach with attention to every detail.
So to encourage the discussion a few points which hit my interest  Wink

In reply #29 you write about lighten the tip plate scavenging them. What I missed was an indication how much you saved. Did you weight before and after?
I try to weight every piece I put my hands on, mainly because I remember well a wing where I lost a few evenings punching holes in all ribs just to find out, at the very end, that it was not worthwhile because the net gain was only very marginal. Or on my actual build putting carbon cap strips only on the last centimetres instead from D-box to TE, difference on the whole wing 0,15g but loss of rib stability.
...
Urs,
I think we have all done that. That is cut lightening holes in wood structure only to find we lost a great deal of strength! I think we see holes in metal structure and think it would be a good idea in our sheet wood parts. As you mentioned the strength in wood is in the grain. I go into a bit of detail here. Again my apologies. (I really need to port over some of my “pearls of wisdom”  Roll Eyes  )
https://www.rcgroups.com/forums/showthread.php?879893-Dumas-spar-in-their-Electric-R-C-kits
I think I can count the destruction of that type of spar as one of my modeling accomplishments!

Now I do think that making ribs like a warren truss, thinning the sheet parts to have a web or making laminated outlines is very advantageous. It was this last idea of laminating the wing tips that got me in trouble with designer. But he did propose the thinning of the tip plate to make the web. Don didn’t think it was worth the effort. And as you noted the benefit was minimal on a 400 gram model.

All the best,
Konrad
Logged

Cut it twice and it's still too short!
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #59 on: December 27, 2017, 11:20:33 PM »

Made time after the holiday madness to get back to glueing some fine laser cut parts. I’m really liking this kit.

I found it a bit odd that I had to make two trim cuts (red lines) to fit the center panel, top end pieces. Normally these 1/64 plywood sheet parts are supplied on size or only need one side (end) trimmed, prior to glueing.

I had to center the lattice to match the tip panels. Also make your cuts so that you have enough waste to be used as sheeting to fill the area adjacent to the spoiler.

Also note that the lattice is off center spanwise. This is to allow good contact area for the panel sealing tape.
Attached files Thumbnail(s):
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Logged

Cut it twice and it's still too short!
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #60 on: December 30, 2017, 06:01:04 PM »

It was very difficult but I think I found 0.8 grams of dead weight in my Chrysalis.

I hate to admit it but I tend to loose small parts very easily. Knowing this I’ve chosen to glue the small carbon wing joiner rods to the center section to help keep all the parts together!  The added benefit is that with one end of the joiner system glued down there is no clearance slop. This eliminates half the slop! With the rods glued to the ribs I can remove the carbon rod balsa stops saving 0.5 grams. Shortening the carbon joiner rods will save 0.3 grams. You might recall that I’m about 0.6 grams in the red with the addition of the diagonal and gussets. I can honestly say that my attention to detail will have removed about 0.25 grams from the as designed structure.  This should easily allow me to end up in the winner’s circle!  Roll Eyes

All kidding aside, I think I have upped the speed at which flutter will occur and at the same time removed some weight from the structure. I think this meets the definition of  an improvement. Now I did damage the design by sanding down the sag compensation. But by the cling wrap test I think I’ve recovered from that error.

Now on to the spoiler. Did I say I hate spoilers?

All the best,
Konrad
Attached files Thumbnail(s):
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Logged

Cut it twice and it's still too short!
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #61 on: December 31, 2017, 12:27:45 PM »

MORE ON SAG COMPENSATION !

...A trick Don told me to see if the sag compensation is working is to “candle” the surface. That is place a straight edge above the wing and look at its reflection in the covering. There should be no kinks or discontinuity in the reflexed line. The more of the wing that shows a smooth reflection the better. I will use this trick to see how well I’ve recovered from my error sanding the sag compensation humps off of a few ribs I damaged.
I should have mentioned that one is looking  chord-wise at the wing when candling the airfoil.

Using Lincoln’s cling wrap it really is interesting to see this sag compensation at work. Yes, there is some distortion at the rib and about 6mm ether side of the rib.  This distortion is not sharp like what we normally see as the covering goes over a spar or turbulator, rather just two gentle bumps.  The rest of the wing area has what looks like a smooth uninterrupted surface. The curve line made by the straight edge reflection does change a bit  as we move it span-wise, but has no sharp changes.

Please remember that I had damaged the ribs. So I can’t comment as to how well the profile of the rib designed by Don Stackhouse really perform at controlling the covering sag to maintain the designed airfoil’s cross section (Please be aware that the rib and the airfoil shape are NOT the same thing in this design). But, the idea of sag compensation looks to be very valid. Also note that Don does not advertise what airfoil section he is using.

While I was able to get the cling wrap to work it was rather difficult as it doesn’t want to cling to raw wood. Next time I’ll try Doculam as a cheep covering to test the prep of the wing prior to covering with Oracover lite or R.A. Microlight films. I did not use Doculam as I feared the high heat would damage the Ducro glue I was using to build up the damaged rib profiles.
Logged

Cut it twice and it's still too short!
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #62 on: January 01, 2018, 09:43:28 PM »

MORE BACK PEDDLING!!

I’ve learned why that model I flew into flutter did not disintegrate! DJAerotech has you place a lot of glass fibers across the dihedral joint. This goes a long way towards making a strong joint, particularly with the “S” glass tying the mid span and tip carbon spars together. Not sure why DJAerotech has you make a rear “S” glass web but it is shown and mentioned in the manual. I have some concerns with my “S” glass as it looks like the CA did not saturate the glass prior to curing.  I say this as it still looks a bit white. Usually the glass will mimic the underlayment color (Black from the carbon) when saturated. Well, this is my experience with epoxy and “E” glass.

Earlier in this thread I said I don’t like glass fibers across the LE joint as one need to feather and fill the weave for a nice covering job. But as DJAerotech has you add fibers across not only the LE but also the carbon spar joints and TE there is no reason not to add the glass to the LE.

This brings me to my back peddling. With the glass cloth across the leading edge there is NO reason to add the leading edge gussets I added. All that CA saturated glass will hold the LE dowel and support the glue joint a lot more efficiently than those gussets. Don’t add them they are just dead weight. I still like the extra diagonal I added to stiffen the wing around the dihedral break.

Now working with composite fibers you don’t want to cut those that span the joints. But you do want to feather them into the adjacent material. To protect the wood and carbon spars mask them prior to feather sanding the glass. I know I’m just about finished when I cut through the masking tape. Now I do sand the whole surface of the glass to knock off any high spots and fuzz. I go back over the glass with thin CA to fill any low spots. Sand and repeat for a nice smooth surface for the covering. Not sure what DJAerotech wants us to do with the dimensional offset the added glass presents to the covering. I just added some more Duco cement to get a nice glass the rib transition.
Attached files Thumbnail(s):
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Logged

Cut it twice and it's still too short!
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #63 on: January 02, 2018, 10:56:39 AM »

As a result of the constraints on most laser cutters, parts are cut only as 2D (X and Y axis). I’m always impressed when I see a laser kit manufacture move into the third dimension. Joe at DJAerotech has done this with the shaping of the elevator leading and trailing edges. This really cuts down on the amount of work required. I’m sure it is a pain, I know if I was doing the shaping I’d be missing most of my fingertips to the sander. Nice touch Joe. Joe is the “J” in DJAerotech. My understanding is that Joe does the heavy work assembling the kits for sale.

The manual does mention this but I’d also like to mention it. It is best to make a “fish bone” with the leading edge of the elevator prior to fitting the elevator ribs to the trailing edge. Conversely it is best to make a “fish bone” with the trailing edge of the stabilizer prior to fitting the leading edge. Now that I haven’t made friends with CA glue and balsa I like to use Titebond to glue the small elevator ribs to the Trailing Edge. I find Titebond helps a lot when it come to finish sanding the elevator.
Attached files Thumbnail(s):
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
« Last Edit: January 02, 2018, 11:06:59 AM by Konrad » Logged

Cut it twice and it's still too short!
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #64 on: January 03, 2018, 03:09:11 PM »

I can’t stress this enough, I’m really impressed with the fact that DJAerotech is engaged in continual process improvement of their product. So much so that in my opinion the plans are due for an update.

First the Center of gravity need to be corrected along with the first dihedral blocking dimension. I’d like to see the triangle assembly aid for the pushrod tubes be removed from the drawings (after all this is a high performance ship not a first time build ship). Also the stab drawing needs to be update with the correct configuration and material callouts. The drawing should show the tongue and mortice feature out at the tip on the leading edge and spar. (I spent too much time looking for basswood parts when the updated elevator is using balsa end pieces). 

I’d like the manual to mention the spar detail at the root (45° and outside radius).
Attached files Thumbnail(s):
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Logged

Cut it twice and it's still too short!
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #65 on: January 06, 2018, 08:02:42 PM »

It works!
It appears that Duco cement works for repairing the damaged rib profile.

I’ve been suffering a lot of hanger rash with the thin delicate wing tips. You might recall that with my modification to the tip the grain is oriented the wrong way. So out of frustration I decided to cover the wing tip. The covering film will add a lot of support to the balsa across the grain.

This is the panel were I damaged rib profile, as a result of inadequate documentation, I sanded off the DJAerotech sag compensation. I’ve tried to repair these ribs by building up the lost humps that make up the covering sag compensation with several layers of Duco cement.  Well, much to my surprise the Duco cement has held up to the heat of the iron.

I hope you can see that there is a low spot in the covering where the rib and spar intersect. This is the extent of the gross distortion of the covering. The rest of the wing does not show any kinks or sudden change in the chord profile (features that will trip the chordwise air flow). It is interesting that while the covering is attached to the spar there is no distortion (kink) in the surface between the ribs. Well, little or no distortion 8mm past the ribs.

Again I need to emphasize that what I’m showing is NOT the result of the as manufactured rib profile. As Don said I destroyed his hard work when I sanded off the sag compensation bumps. Now I still have wing panels where the as manufactured profiles are still evident. It will be interesting to see how much different the covering responds to the as manufactured verses the repaired profiles.

This exercise has shown a problem I found with the laser cutting. I’ll need to make a drawing of this, but due to the laser kerf being tapered the carbon spars are not flat (square) to the assembly aid.
Attached files Thumbnail(s):
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Logged

Cut it twice and it's still too short!
lincoln
Platinum Member
******

Kudos: 29
Offline Offline

United States United States

Posts: 1,752



Ignore
« Reply #66 on: January 07, 2018, 05:25:38 AM »

Build 3 or 4 EZB's and you'll find you're no longer breaking the amazingly heavy, clunky wing tips the Chrysalis Lite is saddled with. ;-)

https://aerofred.com/details.php?image_id=78796
Logged
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #67 on: January 07, 2018, 11:43:53 AM »

Lincoln,
You lost me with that statement. The EZB's wings are hoops, are they not?
Laminated hoops were one of my first idea to eliminate weight. I'm sure they will work as long as one keeps the speed down. But as the flight envelope requires a hi start launch, flutter is a real concern. I learned of this first hand.

Now the issue with my wing tip it that I miss placed the tip plate. To keep from adding tip extra tip camber I added some wood to allow me to sand out the camber. This gave me a nice up turned tip but the way I added the extra wood has it's wood grain going the wrong way for about 8mm. I have pictures of this posted earlier.  

Now I've built a few Penney planes and can tell you that keeping the kind of structure you link to stable and rigid is a challenge! Even when my Penney planes weigh a quarter the structure is very difficult to keep in trim.

The F3-RES Chrysalis wing tip as designed, looks to be one of the few that is integral with the wing's structure. Most of the other F3-RES kit wing tips do look like clunky add ons. Few kits have any of the load bearing members actually attach, over lap, the tip plate. Many kits have the load bearing members stop at the last rib and just glue on the wing tip as an afterthought.

I need to stress that the hanger rash was pretty much limited to the wood I added. Laminating this wood with film both on the top and bottom has given the wing tip massive support. I should no longer have issues with the wing tip during construction of the wing. And I suspect that  in service  my wing tip will be just as or close to being as durable as the one originally designed. And with the added benefit of the webbing feature, recommended by DJAerotech, even lighter than the one found in the kit.

Logged

Cut it twice and it's still too short!
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #68 on: January 08, 2018, 07:30:12 PM »

What I claim got me in trouble with maintaining the as designed rib profile can be traced to the nature of the laser cutter which results in cut sides that are not parallel.

A laser beam has a taper as a function of the focusing lens blue lines. This isn’t too steep, but the resulting heat radiation red lines makes the top of the cut wider than the bottom of the cut. Using feeler gauges I measure a little less than 0.1 mm per cut side.  This means that the shear web (assembly aid) will place the carbon spars as shown. This will result in an assembly that will have a mis-match with the rib and the spar, I call this a castellation. Now normally I address these build discontinuities by using a sanding block to true up any build (cut) errors.  DO NOT sand the ribs with a sanding block trying to get a smooth transition from the rib to the spar! You are likely to do what I did and remove the sag compensation feature of the rib.

I’ve learned that DJAerotech does not perform its own laser cutting. This service is done by competitive bidding. This might be a good thing as it means DJAerotech can out source the cutting to the shop that has the best tool for the job. I would think that cutting the shear web might best be done with a shop bot rooter a straight carbide bit should produce a cut that is less tapered than a laser.

Now there are machine processes that can be used on the laser to minimize this laser kerf taper. As counter intuitive as it may sound using more laser power combined with a higher pulse frequency* and faster traverse speeds will result in the radiant heat cutting less of a taper.  Most cutting houses don’t like to do this as it puts a greater load on the laser gas tube and the positioning cogs need to be more robust to keep the tool path accurate.

*A laser uses pulsed light to cut.
Attached files Thumbnail(s):
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Logged

Cut it twice and it's still too short!
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #69 on: January 08, 2018, 08:46:31 PM »

Don got back to me on my covering of the wing. While Don expects the covering to be bound to the spar as part of the covering process and to aid in making a more rigid panel. He did not expect that the covering would be forced down onto the whole spar. He expects that the area over the rib and adjacent to the rib to have the covering floating a bit. This should minimize the local distortion even more. Please see the detail photo of Don’s (the designer’s) wing.
Attached files Thumbnail(s):
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Logged

Cut it twice and it's still too short!
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #70 on: January 09, 2018, 09:28:09 PM »

Don had asked that I monitor the threads on RCG as that is where, for better or worse, he likes to post general answers.

I that vein I hope Don takes a glance here at HIP.

Please do NOT do what this poster suggests.
https://www.rcgroups.com/forums/showpost.php?p=38956442&postcount=413

For aerodynamic reasons there isn’t much balsa in the sides of the fuselage. Cutting the few balsa fibers there are to allow for an assembly aid (former tabs) extracts far too high a cost in the way of strength. A major requirement of the F3-RES class of ship is the landing points. While lawn darting is frowned upon the fuselage will still see rather abrupt landing compared to the typical sport (Sunday) flier.  The fuse sides are as they are for strength and has nothing to do with a Toledo finish.

As of today the kit has more than ample detent notches in the 1/64 doublers I found the structure more than adequate to self align and hold itself together while one poured thin CA into the joints.

Again I found that one needed to follow the manual and feed the tail boom from the front. The boom is tapered and sliding against the taper makes it all but impossible and still have a tight carbon boom to bulkhead joint. (Do make sure you aren’t pre loading the tail boom nullifying the offset in fixture “F5” Roll Eyes ).
« Last Edit: January 09, 2018, 09:42:23 PM by Konrad » Logged

Cut it twice and it's still too short!
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #71 on: January 12, 2018, 06:58:02 PM »

This post is more of a shout out to Dan Stackhouse (DJAerotech). I’d like to thank Don for his great explanation on how to film cover an open bay wing that has thin trailing edges.  Far too often us engineers are accused of over engineering things (going into too much detail). But this process works great, I tried to post the process earlier in this thread.

I’ve been covering wings with iron on films (MoneyKote) for 40 years and always dreaded covering thin trailing edges. This process has worked better than I thought it would. In one of the photos I’m showing that the trailing edge did not distort (bend up of down) as I shrunk the film. You can also see in the top view that the ribs are straight and did not wander as the film was shrunk.

I know Don and Joe don’t want to put too much info in the manual, as far too many folks see a thick manual and pass out from information over load. But in this case I think DJAerotech should add an addendum to their manual describing what Don wrote to me early in the build process.  Don has already gone through the effort to write this up. I think all his customers should be aware of how the film cover a thin trailing edge. If it would help I can supply Joe with the vector drawing I used in my post or I could clean them up and offer them in a JPEG format for ease of publication.

Again a great set of instructions and a fantastic out come.

Thanks DJAerotech!

All the best,
Konrad
Attached files Thumbnail(s):
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Logged

Cut it twice and it's still too short!
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #72 on: January 21, 2018, 05:21:27 PM »

Starting to train the parts to fly in close formation. The wife thinks I’m nuts as I’m starting to make swooshing sounds while holding the model. Undecided

I’m happy to report that the Duco cement repairs of the rib profile has worked great. I had feared that much of this build up would melt when I applied the covering. This wasn’t the case when attaching the Oracover lite (Hanger 9 Park-lite) at 105°C. I think I also used this low heat setting to shrink the covering.

To my eye, the covering is responding (sagging) the same for both the as cut and repaired ribs. Time will tell if I notice and difference in the stall and speed performance between the as cut rib and repaired rib panels during actual flight.  
Attached files Thumbnail(s):
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Logged

Cut it twice and it's still too short!
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #73 on: February 04, 2018, 08:24:51 PM »

If I haven’t said it before I’ll say it again “I hate spoilers”.

With that out of the way I decided to use a different method of control for the spoilers. I’ve had very pool luck with magnets, springs, and strings. The method I’ve chosen has the servo control both ends of the spoiler travel and it does this without needing to “size” the push rod to perfection. It is self aligning at full open and fully closed through the radio.

The photo is showing the initial set up. I will need to move the output shaft of the servo closer to the hinge line of the spoiler. This means I’ll have to cut off the front servo mounting tab and raise the servo 2.5 mm higher than shown. The push rod is held only by the connector on the spoiler. The others on the servo arm are just a slip fit allowing the rod to slide as needed, compensating for the fact that the rotational axis of the hinge and the spoiler are not aligned.

A nice feature not mentioned is that the spoiler is a composite made of different density balsa. The last third (the very thin side) is made from harder balsa than the thicker side of the spoiler. Nice touch from DJAerotech. Please cover the spoiler using much the same method described earlier for covering the trailing edge

To control the spoiler’s down position at the ends I’ve added some stops.
Attached files Thumbnail(s):
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
Re: Chrysalis F3-RES, DJAerotech I'll be improving it until it doesn't work
« Last Edit: February 04, 2018, 09:46:12 PM by Konrad » Logged

Cut it twice and it's still too short!
Konrad
Platinum Member
******

Kudos: 35
Offline Offline

United States United States

Posts: 1,941


Topic starter
Measure twice cut once



Ignore
« Reply #74 on: February 06, 2018, 07:26:45 PM »

While fitting the spoiler servo I ran into a minor issue with the kit. The cut outs to allow the spoiler servo leads will NOT allow the Jr style connector on the D4.7 servo to pass. One can ether un-pin the connector and thread the lead through the two holes (one in the center rib the other on the plywood center sheeting). Or open up the holes 0.35mm. I chose to open up the holes about 0.015".
Logged

Cut it twice and it's still too short!
Pages: 1 2 [3] 4 5   Go Up
Print
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Valid XHTML 1.0! Valid CSS!