Date: Mon, 1 Dec 1997 18:32:50 -0500 (EST) From: Nigel Field Subject: Re: COZY: Cosy operating Costs At 11:10 AM 12/1/97 -0600, Carl wrote: > > >Propellers: Yes, I have considerable wood working experience, but by the >time I would study the process, buy the necessary equipment, and then my >time is valuable also. I have elected to purchase mine. Prehaps Nigel can >write a bit on whats required to make your own propeller. (Equipment, >costs, sources of material, how the shape is arrived at, how many props >made before happy with performance. > > Good post Carl, thanks. We tend to work on a tighter budget up here where our dollar is now about 2/3 of yours. Still we get by and have a ton of fun. And I was not kidding my costs are very low in comparison. Re props. I think Marc would not appreciate a pile of discussion on prop design and fabrication here on the list. I have been corresponding privately with 2 other list members on how do do it though. For background I designed and made 2, my first a 3 blade and second a 2 blade. This was more from neccesity than economics as I use a high RPM engine and homemade gearbox so prop selection was limited. Both perform very well with the 3 blade clearly in the lead by all respects, ie climb, cruise and smoothness. They are E glass composite over a soft pine core with thin carbon tips and integral spars to take half the centrifugal blade load. The aerodynamic design is done with a SW application called Prop optimizer and the structural stress calcs are done in Mathcad with a simplified Lotus version as well. The designed safety factor is 5 to 1 for both at 9 lbs weight 3200 RPM. They have never been tested to destruction but my 3 blade has over 100 hrs on it and it looks and flies like new. The 2 blade 1 hr, still in primer. Both cost about $60.00 each for wood, glass, carbon, resin, resorcenol glue, most of which I (and any Cozy builder) already has. Tools are a spoke shave, angle protractor, drill press, hacksaw, chisel, couple spade drill bits, home made clamps from 2 X 4 pieces and threaded rod, plus usual composite stuff we all have. Your time, well its free if you build your own airplane. About 50 hours should do it and thats a SWAG which will vary. If anyone is really serious about trying it then perhaps I can help. I could provide the SS and/or mathcad file for stress calcs and run the numbers for the blade profiles. The SW needs an acurate engine HP curve plus airplane drag to give accurate results. But I have most of that already and can probably get pretty close on the drag especially if flight data is available. The construction description is long and currently consists of about 20 Emails and a fax diagram to others with lots of good questions and the answers. I'm not really set-up to write a detailed article. But if there is enough interest perhaps we could look at creation of a temporary list server just for this. I have no idea how to guage this, it may be a bust idea. I just do my thing up here and have a bunch of fun making stuff, but have no objection to passing on anything learned. But my 3 blade is the best prop I have ever flown and I've tried many props. (no I'm not trolling) Perhaps the other 2 could let me know how they are progressing so we can see if its a feasible effort. Ian ya go to stop voluntering me for stuff. Nigel Field (Cozy III on hold again) From: Lee Devlin Subject: COZY: Nigelprops (making your own) Date: Wed, 3 Dec 97 16:26:55 MST Nigel wrote: > I have no idea how to guage this, it may be a bust idea. I just do my thing > up here and have a bunch of fun making stuff, but have no objection to > passing on anything learned. But my 3 blade is the best prop I have ever > flown and I've tried many props. (no I'm not trolling) Perhaps the other 2 > could let me know how they are progressing so we can see if its a feasible > effort. Nigel has graciously provided me with information on how to build a prop for my LongEZ. He's save the email exchanges so maybe Marc can put them in a big file and use the list server to give this information to anyone wishing see what's involved. I bought the materials and tools I will need but due to my travel schedule I haven't had a chance to start the prop yet. It looks like about the same amount of work one might spend making a canard. Since the price of a 3-blade prop is around $1700, I think it's a good trade off of time for money. Perhaps Nigel can run the numbers for an O-360 Cozy MKIV and post it which would satisfy most builders and avoid having a lot of individual requests for the data. I have laid out my LongEZ prop templates on a cad system and plotted them out full scale for each of the 9 stations per blade. I also plotted the planform for the blade. I'm hoping on spending some time over Christmas making a prop. I'll keep this list updated on my progress. Lee Devlin Date: Wed, 3 Dec 1997 22:26:03 -0500 (EST) From: Nigel Field Subject: COZY: Prop Making Hi all, Just a short note on the status of prop making. We currently have 8 folks who are interested in building. Marc Z has kindly offered to host the existing files on his server for download. He also suggested we conduct open discussions on the Cozy list as needed, but if the bandwidth gets too high we may look at another option. Will clearly title all prop stuff so those not interested can hit the delete key. Lee Devlin has also kindly offered to allow all of our Email correspondance be made public to the list. Perhaps we can ask him to share his CAD files also for those who are into that. I have forwarded some files to Marc for review and will get the rest to him within a few days. Getting some diagrams scanned at work and will take a few more pics to add to what I have. I will run numbers for anyone who wants them and send the files privately as they are large, but will archive a popular one ie MK IV with 0-360 for reference. So in summary it should all be there in a short time for those who want to make their own props. You guys who want to play need to get a spoke shave and honing stone for Xmas. Nigel From: Marc J. Zeitlin Subject: COZY: Propeller Fabrication Date: Mon, 8 Dec 97 17:03:11 EST People; Pursuant to Nigel Field's previous discussions regarding Propeller Fabrication, I have placed his correspondences with Lee Devlin (with both of their permissions) on the mailing list server for retrieval by those who might be interested in fabricating their own propellers, or at least learning about the process. Send an email to: majordomo@hpwarhw.an.hp.com with the lines: get cozy_builders prop_fab end in it. This will email you the file. Happy reading and propeller construction - this looks like a relatively easy way to save $1500 and learn a lot. -- Marc J. Zeitlin Email: marcz@an.hp.com From: "Fred I. Mahan" Subject: COZY: Propeller Fabrication Date: Tue, 9 Dec 1997 14:59:07 -0500 Speaking of propeller fabrication, I recently received an interesting mailing from J.B. Allred of Allred & Associates, Inc. It's for a "GemInI (that's Gemini with two capital i's, not L's) Propeller Duplicator." The device is a bed or framework in which a prop and a prop blank are placed side by side. The framework has a moving crosspiece on which is mounted a pantograph. One "finger" of the pantograph rests on the prop, and the other "finger" of the pantograph has a small router. You move the finger over the prop and the other routs another propeller in the same shape. I have no connection with this company J. B. Allred Allred & Associates, Inc. J-E Professional Building 5566 Jordan Road Elbridge, NY 13060 Tel: 315-689-1626 Fax: 315-689-1438 No e-mail address or web site listed Fred in Florida Long-EZ N86LE Defiant project From: lschuler@cellular.uscc.com Date: Wed, 17 Dec 97 17:49:17 -0600 Subject: Re[4]: MORE PROP Stuff I wrote: >snip >>transition from hub to blade. >>snip >> >>Larry You wrote: >OK understand now. This is hard to describe. The blades make a natural >transition into the flat hub which has some curves but pretty much ends >flat near the hub. The glass will lay up fine over these curves. Have >attached a photo of the prop on my VE which shows this. Double click on it >and it should launch a photo viewer in Win 95 or use whatever else you have >to view a photo. >Nigel Gotcha... much thinner than, say, a Performance 3-blade for IO-360. Ain't glass neat! Basically just make a nice transition and copy for all blades.... simple enough. BTW: (humorous note) noticed in text that blades are cut +/- 30 degrees, then butted to make the hub; however, text doesn't say that the 30 deg is off the perpendicular to blade.... I kept adding up ((30 + 30) * 3) = 180 assuming the angles were taken off the blade axis (which would be, in fact, +/- 60).... Dawned on me this morning while sitting in the library (head) that you refer to the perpendicular even though not stated! Get my best thoughts in the library; have no idea if it's due to lighter load or not; wife tells me once in a while that my brain is in the wrong place, but heart is in the right place..... :-) Larry [Neurons are just pesky little gremlins that organize themselves once in a while to do something that one neuron alone cannot do....., but mostly just mischievous, fun stuff like forcing one's arm to lift a beer; or, writing something like this.] From: lschuler@cellular.uscc.com Date: Wed, 17 Dec 97 18:02:13 -0600 Subject: Re[2]: Prop Group Communications Marc wrote: >... you can group reply to either this or Nigel's message, and just the ten of us will get it. Nice offer to just use Cozy list. How about if we just use alias list for all prop Q & As? But as you say would lose the archive capability unless Nigel can save it all, but that's more work.... BTW: Nigel, really super work and very gracious of you to share. Larry Date: Wed, 17 Dec 1997 23:43:22 -0500 (EST) From: Nigel Field Subject: Re: Prop At 03:11 PM 12/17/97 -0500, Nick Ugolini wrote: >Thanks for the reponse. > >I will drop you an email tomarrow (Thur) with my fax number. Boy, I cant >tell you how excited I am about building my own prop (3 bladed). I think it >is all about having something very few others have (like our >planes)....besides saving lots of $$ > > The main reason I wanted it is because of the 3 >different plane(engines) I have. >I am fixing the Varieze up to sell (needs a 2 blade), and will build a 3 >blade for the long, and a 3 and 2 blade for the cozy. (one to use, and one >ready for emergency shipping). Wow you got quite a stable there Nick. Have sent a .JPG of the prop diagram separately. I read it best with Internet Explorer, gives full size. Anyone else need the stress calc Spread Sheet or the diagram, just ask, can Email or fax diagram. >Did you use S-glass roving for the spars, or did you use carbon? I wanted >to get an order together, and the only things I believe I will need is the >3/4" tube, the roving, and some carbon glass (what type would you suggest >Bid, UID?). I can get the wood locally, have BID and UID, and the epoxy. I used E glass tows from spar tape same as used in the LE and Cozy wing spars. Cut a piece a little longer than needed then separate it into the tows. Cut them shorter and shorter as you fill the troughs until just protruding above the blade, then after cure sand them down flush. This gives a good bond to the skin with no dips of joggles. The carbon is part # 5-GFT from Wicks, get enough for 18 plies 5 inch long plus a bit more, say 10 ft. I used a solid piece of .75 inch alum round stock for the centre spar support insert and bored it 3/16 on the lathe, but you could use thick walled pipe. If you can't find any with .25 wall thickness you could substitute 4130 pipe, about .068 should be OK. Just protect the inside so it wont rust and fail. The spar loads are pretty high so thin walled alum might not hack the loads. This question raises an important point of mixed modulus materials. Carbon is far stiffer, a higher modulus (stretch) than glass. If you take a strand of steel wire with a breaking strength of 20 lbs and you put it in parallel with a rubber band also with a 20 lb break strength, and then stress them, you don't get 40 lbs tensile strength. The steel will take 99.9% of the load until it fails at 20 lbs then the rubber will see all the load and fail also. The same happens with carbon and glass used together in the same load path as is the case here. So all glass, or all carbon, but not mixed together. I know I use carbon on the tips but that is for stiffness and to keep it thin. The load is transferred into the glass skins which overlap the carbon, so the carbon becomes just very stiff core material. > >I have the boss on the prop extention, so I will probably machine a plate, >and use longer prop bolts. What do you think of this idea to accomodate the prop extention boss?: Drill >a hole in the hub (for the 3/4" tube) and on the other side of the prop >(extention side) drill a 2"? (whatever the size of the boss is) hole and >insert a 2"tube in it to wrap the roving around.. IE, a 3/4" tube 1/2 of the >way through the prop, and a 2" tube 1/2 of the way through the prop. the >wood and glass should keep the tube from shifting, and the tension forces >should be equally distributed no matter what the size of the tube is. It >just will not be a continuous tube (to accomidate the boss). Yea I discussed that idea with Lee but I don't think the spar tows will clear the bolt holes this way. I would machine a plate. > Or maybe use a >2" tube all the way through the center of the hub. I think it might be >better than wraping the roving around the prop bolts..and easier to make >than the offset plate (and no bolts to buy). I agree, I don't like the idea of looping around the prop bolts much either, it would require alum inserts be built into the hub and drilled through after skinning, but it would probably work. I really think the best approach is to use an aluminum spacer to clear the boss. I had completely forgot about the centering boss when I made mine as I don't use one on my hub. The bolts centre it fine. I am sure we can get some alum spacers made and if someone is really, really stuck I can turn one up on my lathe from a slice of round stock. Or just machine off the boss on the alum extension, thats the easiest way. Hope this covered it. Nigel From: Marc J. Zeitlin Subject: Re: Re[2]: Prop Group Communications Date: Thu, 18 Dec 97 9:03:45 EST Larry Schuler wrote: > Marc wrote: >>... you can group reply to either this or Nigel's message, and just the > ten of us will get it. > How about if we just use alias list for all prop Q & As? But as you say > would lose the archive capability unless Nigel can save it all, but that's > more work.... Since people seem a bit reluctant to post to the COZY list, here's what I'll do: Since I'm in this prop fab group and get all the email (at least the ones that are sent to all 10 people, or ones I'm copied on) I'll save all the posts into the COZY archive as if they were sent to the full COZY list. This way, they'll be retrievable by anyone on the COZY list (including the 10 of us) but not by anyone else. They will NOT be in the monthly archives unless they're sent to the COZY list as a whole. Anyway, to retrieve the prop fabrication discussions via majordomo, send a message to: majordomo@hpwarhw.an.hp.com with the lines: get cozy_builders prop_fab get cozy_builders topics/prop_fab.txt end After January 1st, 1998, you'll want to add: get cozy_builders topics97/prop_fab.txt to that list of files to get, at least the first time :-). This will retrieve the original "prop_fab" file that most of you already have (so you could leave that first line out, if you don't want to retrieve it again) and will also get the current version of the "prop_fab" discussion archives that I'll be saving. If you don't want a message saved to the archive, either don't send it to me :-), or mention specifically in the message that it's NOT for archiving. -- Marc J. Zeitlin Email: marcz@an.hp.com From: Marc J. Zeitlin Subject: Re: Re[2]: Prop Group Communications Date: Thu, 18 Dec 97 15:09:09 EST People; Since I was so vociferous in my statements regarding the T.E. closeout issue on the canard that one COZY list member had, I was asked about this issue with the props, too. The reason is that Nigel has called out a glass-flox-glass T.E., with essentially no G/G bond at all. Nigel says: "The TE is relieved of wood about .2 inch in for a flox bead full span, then skin the top same as bottom." Now, Nigel has been flying this prop for 150? hours or so, so it's obviously not going to explode :-), but..... Here was my take on this issue to the questioner - Nigel (and anyone else), please let us know what you think of this evaluation. > ...... Just thought I'd mention Nigel's > trailing edge close-out using flox between top and bottom skins of the > prop.... Seems a prop would be taking much more of a beating in > torsion than a canard, especially in the wake of the strakes???? Your logic is correct here, and we can intuit one of three things (or maybe more): 1) Nigel will have problems due to fatigue of this joint in the future. 2) It doesn't make any difference how strong this joint is. 3) I was being too conservative in my exhortations re: the wing trailing edge, and there really won't be a problem either there or on the prop even though it might be a bit less strong or stiff. My guess is #3 :-), _but_ we've got to remember that Nigel's running a EA-81 at ~5000 RPM with 100 HP, not a 0-360 180 HP at 2700 RPM. Lots more pounding from the 0-360, and lot's more thrust. I would lean toward closing out the trailing edge of the prop just like the wing, but with a flox fill rather than micro - this would put the flox on the OUTSIDE of the glass, rather than between the glass layers. Maybe have ONE layer of glass over the flox to safeguard against fatigue cracks. What say you all? -- Marc J. Zeitlin Email: marcz@an.hp.com Date: Thu, 18 Dec 1997 17:39:13 -0500 (EST) From: Nigel Field Subject: Re: Re[2]: Prop TE closeout At 03:09 PM 12/18/97 EST, Marc Z. wrote: >People; > >Since I was so vociferous in my statements regarding the T.E. closeout >issue on the canard that one COZY list member had, I was asked about this >issue with the props, too. The reason is that Nigel has called out a >glass-flox-glass T.E., with essentially no G/G bond at all. > >Nigel says: > > "The TE is relieved of wood about .2 inch in for a flox bead full > span, then skin the top same as bottom." > > >Now, Nigel has been flying this prop for 150? hours or so, so it's >obviously not going to explode :-), but..... > >Here was my take on this issue to the questioner - Nigel (and anyone >else), please let us know what you think of this evaluation. > >> ...... Just thought I'd mention Nigel's >> trailing edge close-out using flox between top and bottom skins of the >> prop.... Seems a prop would be taking much more of a beating in >> torsion than a canard, especially in the wake of the strakes???? > >Your logic is correct here, and we can intuit one of three things (or >maybe more): > > 1) Nigel will have problems due to fatigue of this joint in the future. > > 2) It doesn't make any difference how strong this joint is. > > 3) I was being too conservative in my exhortations re: the wing > trailing edge, and there really won't be a problem either there or > on the prop even though it might be a bit less strong or stiff. > >My guess is #3 :-), _but_ we've got to remember that Nigel's running a >EA-81 at ~5000 RPM with 100 HP, not a 0-360 180 HP at 2700 RPM. Lots >more pounding from the 0-360, and lot's more thrust. > >I would lean toward closing out the trailing edge of the prop just like >the wing, but with a flox fill rather than micro - this would put the >flox on the OUTSIDE of the glass, rather than between the glass layers. >Maybe have ONE layer of glass over the flox to safeguard against fatigue >cracks. > >What say you all? > Here's my take on it. I think it is a very valid concern and Marc is quite correct about 0-360 torque pounding. My prop runs at 3200 RPM but gets very smooth torque as power is passed through a torsional isolator before it gets to the gear box and the EA-81 only makes little pulses, just a lot more of them per second. The design as it currently exists is for a 100 HP engine and will work OK on an 0-235 even with the bigger power pulses. I realize Cozy guys need a beefier prop for the 0-360 or even the Franklin 6 cyl. I had planned to discuss all that and work through the changes with all the brains contributing. So lets take a first shot at it. First the spread sheet can be used to work the stresses and derive the spar sizes and skin thickness for any desired safety factor (sf) for 2, 3 or more blades. Mine is set at sf 5 and I think thats a good conservative number to shoot for. Dropping it to say 3, would only save a tiny bit of weight so not much point. If you play with it you will notice a big reduction in sf if you go to 2 blades. So better strength and stress distribution can be achievd with 3 blades, besides I think they work better. The weight of an 0-360 prop is an unknown at this time because nobody had made one yet. I would guess at about 12 lbs, vice a measured 9 lbs for both my 100 Hp props. I would suggest increasing the spar dimensions to .3 inch and add one extra skin ply top and bottom full span. If you input this in the SS it gives sf of 4.97. Next I would add another bid ply on both faces of the hub extending about 3 inches onto the blade roots, 3 total per side staggered a little. If each ply is orinted + - 45 deg to the blade radial line this should provide could engine torque transfere into the blades. Don't think it needs much more than that. Now for TE closure. Flox is pretty tough cellulous (sp) fibre with good structural properties. But I think it would be better to close out the TE in the more traditional way with a glass to glass bond first, then fill in the joggle with wet flox and sand it to match the blade contour. Its about the same effort as the wood has to be relieved away anyway, so just put a radius on the top edge with sand paper so the top skin glass can gently curve down to meet the bottom plies. Since the skins are layed up full span first with the length decreasing as plies are added toward the root, it would be more difficult to cover the flox fill with the last ply. This would mean a cure and transition sand of the overlaps, then add another ply full span. This raises another question as to why use flox at all. I think wet micro into the joggle would bond just as well and be lighter, yet still flexible enough. If some did come out on testing and found unsuitable it would be fairly easy to sand it out and replace with flox, but I think it would work fine. There is no reason to use the structural properties of flox just as a filler, and it is a lot heavier. I will make my next one this way and see. Any and all comments welcomed and appreciated. Nigel From: lschuler@cellular.uscc.com Date: Thu, 18 Dec 97 17:17:17 -0600 Subject: Re[4]: Prop Group Communications HI all, As if y'all couldn't tell who raised this nasty little issue... >snip >My guess is #3 :-), _but_ we've got to remember that Nigel's running a >EA-81 at ~5000 RPM with 100 HP, not a 0-360 180 HP at 2700 RPM. Lots >more pounding from the 0-360, and lot's more thrust. Agree somwhat particularly on the ponding a prop would receive from a Lycosaurus vs a super-smooth 230+ HP Subaru SVX (that's another raging debate); but, here's another point on the joggle vs flox in this particular application: The glass here is totally UNI with fibers running precicely parallel to the TE (which is parallel to the blade radial). Not so with canard; which has torsional BID fibers laid in with the UNI. >I would lean toward closing out the trailing edge of the prop just like >the wing, but with a flox fill rather than micro - this would put the >flox on the OUTSIDE of the glass, rather than between the glass layers. >Maybe have ONE layer of glass over the flox to safeguard against >fatigue cracks. Ok, Marc; here we go again..... Per your previous description of the old soda can and mine of the hinge (see archives): with no BID (or core) here and UNI fibers parallel to TE, the joggle in your soda can has been cut, or at least made extreemely thin. The gap between UNI strands is indeed an epoxy hinge now. I say tapering the Flox between glass is quite adequate in "this application". Non-Lycosaurus application. My 0.00 worth. Larry From: Marc J. Zeitlin Subject: Re: Re[4]: Prop Group Communications Date: Fri, 19 Dec 97 8:58:58 EST Larry Schuler wrote: >Ok, Marc; here we go again..... with no BID (or core) >here and UNI fibers parallel to TE, the joggle in your soda can has been >cut, or at least made extreemely thin. The gap between UNI strands is >indeed an epoxy hinge now. I say tapering the Flox between glass is >quite adequate in "this application". This is a good point - I had completely missed the effect of the pure UNI layers on the utility of the G/G bond at the T.E. Assuming no torsional wrap, I'll concede that whether you use G/G or G/F/G would be meaningless (if it isn't in the general case, anyway :-) ). As I mentioned in a previous private email to Larry, I'd still like to investigate the stress levels due to torsion and/or thrust, but assuming that they're low in comparison with the inertial loads (as seems to be the case), then the current design should be fine. It seems, at least in Nigel's current low HP prop on a Subaru, that the torsional, thrust, and/or engine torque impulse stresses are absorbed by the wood core of the prop, while the glass absorbs the inertial stresses and protects the exterior. -- Marc J. Zeitlin Email: marcz@an.hp.com From: Lee Devlin Subject: Re: Re[4]: Prop Group Communications Date: Fri, 19 Dec 97 11:14:43 MST Marc wrote: > It seems, at least in Nigel's current low HP prop on a Subaru, that the > torsional, thrust, and/or engine torque impulse stresses are absorbed by > the wood core of the prop, while the glass absorbs the inertial stresses > and protects the exterior. I haven't run the calculations myself, but at a prop forum (by Global, http://www.eagle.com/global/index.html) at Oshkosh this past summer, the speaker said that the inertial loads are nearly two orders of magnitude greater than the other loads on a prop. Lee Devlin Date: Tue, 23 Dec 97 9:35:04 EST From: "Nick J Ugolini" Subject: Spar Hi, This is my first post to the "prop group" and I am planning to start building my prop next week. I want to publicly thank Nigel for his support and innovative design. All the design drawings will be in AutoCad R14 if anyone is interested in file copies when completed. The first question I want to throw out is spar design. The prototype is a square section ~.25 x.25. Ideally, you would want to maximize the contact area between the spar and the prop sheathing to transfer the loading of the spar. Another possibility would be to make it narrow and deep to increase the spar stiffness. ? If we are only worried about tension loading do we need a trough at all. A grove could be cut in the hub to allowing the fibers to be wound around the shaft and fanned out on the prop surface to distribute the tension loading a bit more equally on the blade. Thought? Date: Tue, 23 Dec 1997 21:00:14 -0500 (EST) From: Nigel Field Subject: Re: Spar Thanks to Marc again for offering his list server experise and facilities. I think it will be better on a list server. Please welcome Cliff Cady to the group, I have added his Email to the list, and also included my own which I forgot to do last time, see below for latest list. Nick Ugolini wrote: >I want to publicly thank Nigel for his support and innovative design. Thanks for the kudos but really its a pleasure to network with other canard builders a little right of centre, or is that off centre, or excentric, well you know, folks who can see past the status quo. Anyone who is willing to make their own prop fits that group. >All the design drawings will be in AutoCad R14 if anyone is interested in >file copies when completed. Thanks Nick, please include me in that list as I now have autocad R14 but still low on the learning curve in making drawings. >The first question I want to throw out is spar design. The prototype is a >square section ~.25 x.25. Ideally, you would want to maximize the contact >area between the spar and the prop sheathing to transfer the loading of the >spar. >Another possibility would be to make it narrow and deep to increase the >spar stiffness. >? If we are only worried about tension loading do we need a trough at all. >A grove could be cut in the hub to allowing the fibers to be wound around >the shaft and fanned out on the prop surface to distribute the tension >loading a bit more equally on the blade. >Thought? Marc wrote: Here, here, on the thanking Nigel part. Your welcome again, but you're doing half the work so thanks to you also. As far as the spar goes, I had similar thought. It wasn't clear to me why to have spars, rather than just more UNI skins, or as Nick suggests, a "fanning out" or distribution of the spar fibers under the skin. Nigel, if you could give your original reasons for the rectangular spar, we'd appreciate it. ......................... Sure, here is the thought process that led to the first one. I originally wanted to make a 3 blade from maple laminates and employ multiple overlaped splices at the hub to hold it together (a ton of work btw), like the commercial ones do. But the price and availability of high grade maple up here in maple leaf country soon canned that idea. So I thought about a soft and light wood core with composite skins, similar to what I had done already to prop tips. The problem was how to keep the hub together without a lot of complexity. Since the wood would not be carrying the centrifugal loads, I reasoned that simple butt joints at the blade roots would work to form the wood core. The loads then needed to be distributed and contained within the hub structure leaving only the engine torque loads transmitted into the hub. Once the blade root stresses were calculated (as in the SS) the idea of using the looped spars and centre insert became clear. It could carry half the loads while keeping the hub together and the skin could take the other half. So thats how it came to be. I have since thought a bit more about better load transition into these spars as discussed above. There is actually a lot of surface area for the skin to transfere this load as the spars go out to 40 % radius. The skin is all uni and layed up span wise so the angle to the spars which make a gentle V shape, is slight, allowing lots of distance for the loads to transfere. Same argument for the skin tows all parallel, to pass loads inboard to the spars. However I do like the idea of making the spars wider and shallower to provide more skin contact area for even better load transfere. I think that due to the centre insert where they all interloop, they should remain square until clear of the hub diameter then could fan out to say .7 inch wide or so as they progress down the blade. A few calculations would determine the trough dimensions to provide sufficient spar glass to maintain a linear taper in cross sectional area. The spars could be assigned stations along their length and the depth cut and measured to conform to this model. Its a little more complicated to do but certainly not a big deal. I think fanning out the spars without grooves would leave an ugly bump and upset the aerodynamic properties, sort of like laying up your wing spars over the foam cores with no troughs then putting the skins on, so I'm not in favour of that one. No doubt it would carry the loads OK and provide good transfere from the skin, but would have to transition into grooves near the hub where they focus anyway. Saves a bunch of micro work on the blades to transition the bump. Naw bad idea IMO. As for skin only and no spars at all, the calcs show it would work OK if enough skin was added to make up for the missing spars, after all the stress is maximum at the skin. But it leaves no formal structure to hold the hub together, no alternative load path to keep stress entirely within the hub. Should you get a object stike on the inner part of a blade like a rock or fuel cap, exhaust clamp or whatever, it could damage the skin and leave no alternative structure. With a safety factor of 5 the skin could be cut right though and the blade will remain on with a safety factor still remaining. Thats a weak argument though. Its really up to the builders but I don't think I will make one that way. I just "feel" more comfortable with the spars in there. Nigel ........................................... Here is the latest list: Phillip.Johnson@kan.lmcda.lmco.com Lee Devlin Stet Elliott Nick Ugolini marcz@an.hp.com lschuler@cellular.uscc.com "norm & monda" CCady gperry@usit.com Ken Reiter dkuechen@ccrf-mail.gtri.gatech.edu nfield@cyberus.ca Date: Tue, 23 Dec 97 11:37:17 EST Subject: Re: Spar People; Please be careful when sending to the "prop_fab" list - Nigel was left out of the receivers list, so never saw Nick's message. This is why I've copied the whole thing here. (I've just set up a small mailing list for the e_panel folks - I think I'll do the same for the prop_fab people as well, to ensure that everyone gets all the messages). Nick Ugolini wrote: >Hi, >This is my first post to the "prop group" and I am planning to start >building my prop next week. >I want to publicly thank Nigel for his support and innovative design. >All the design drawings will be in AutoCad R14 if anyone is interested in >file copies when completed. >The first question I want to throw out is spar design. The prototype is a >square section ~.25 x.25. Ideally, you would want to maximize the contact >area between the spar and the prop sheathing to transfer the loading of the >spar. >Another possibility would be to make it narrow and deep to increase the >spar stiffness. >? If we are only worried about tension loading do we need a trough at all. >A grove could be cut in the hub to allowing the fibers to be wound around >the shaft and fanned out on the prop surface to distribute the tension >loading a bit more equally on the blade. >Thought? Here, here, on the thanking Nigel part. As far as the spar goes, I had similar thought. It wasn't clear to me why to have spars, rather than just more UNI skins, or as Nick suggests, a "fanning out" or distribution of the spar fibers under the skin. Nigel, if you could give your original reasons for the rectangular spar, we'd appreciate it. -- Marc J. Zeitlin Email: marcz@an.hp.com From: Marc J. Zeitlin Subject: PROP-FAB: Test message to PROP-FAB mailing list Date: Wed, 24 Dec 97 11:11:25 EST People (or whatever the 11 other of you would like to be called :-) ): I've set up a mailing list for the prop-fab discussion group. Here's the info on how to use it: To send mail to everyone on the list (to ask a question, answer one, or just comment to all) send mail to: prop_fab@hpwarhw.an.hp.com This will forward your message to all members of the list. I am on the list as well, so you don't have to "cc" me - I'll get the message too. As a matter of fact, if you send mail to the list, you shouldn't "cc" anyone on the list - they'll get two copies! For those of you using CC:MAIL who are responding to a previous post, try to "forward" the message rather than "replying" to it - I believe this will alleviate the dual message problem. If you have any questions, feel free to ask either me (for list specifics) or everyone (for Propeller Fabrication related info). The only archiving of these messages will occur in the cozy archives, under the "prop_fab" topic. They will not be in the monthly archives, but will be accessible via either the web page or the majordomo mail list server. Happy Propeller Fabrication :-). -- Marc J. Zeitlin Email: marcz@an.hp.com From: lschuler@cellular.uscc.com Date: Wed, 24 Dec 97 11:48:47 -0600 Subject: PROP-FAB: Spar Fan I agree on the need for the spars to keep the hub together. Nothing wrong with additional margins.... Here's a thought on fanning (based on the 100 HP design: I believe you currently taper spars to near nothing at 40% of span. Rather than doing that, how about keeping all fibers same length (total of the 40% span plus hub loop) and spread them evenly across the blade beginning at the hub-blade intersection. I'd keep the existing trench the same across the hub and around the center. This will provide the needed spacing for the bolts. My initial calcs for depression needed in the blade shows that the trench/depression beyond the hub would taper from .25" wide and deep at the hub to 2.5" wide and .025" deep at 40% for single tow depth at the 40% mark; OR, if a two-tow depth is desirable (0.05" depth), the width woud be 1.25" per spar. As you metion, a bit more involved in making the trenches, but certainly do-able. The layup schedule for transition from square at the hub to the "fan" would need to be looked at. May not be important beyond the 6" hub diameter. I'd be inclined to lay in the tows until it "looks" good with a bit of overfill and sand to shape after cure. Larry Date: Wed, 24 Dec 1997 14:57:42 -0500 (EST) From: Nigel Field Subject: PROP-FAB: Re: Test message Just a short one here, great to see the list server up and running. Marc I don't know how you find the time. I'm out of here for 3 or 4 days to visit family for Xmas so see y'all in the new year. Hope Santa brings lots of good airplane stuff. Best of the season and happy building to all. Nigel From: Marc J. Zeitlin Subject: PROP-FAB: Re: Spar (fwd) Date: Thu, 25 Dec 97 12:06:15 EST Nigel Field wrote: >However I do like the idea of making the spars wider and shallower to >provide more skin contact area for even better load transfer. I think that >due to the centre insert where they all interloop, they should remain square >until clear of the hub diameter then could fan out to say .7 inch wide or so >as they progress down the blade. A few calculations would determine the >trough dimensions to provide sufficient spar glass to maintain a linear >taper in cross sectional area. The spars could be assigned stations along >their length and the depth cut and measured to conform to this model. Its a >little more complicated to do but certainly not a big deal. I won't commit at this point (don't know when I'll get the time) but I'd really like to do a Finite Element Analysis of this spar/skin structure with some varying parameters. I'm having a difficult time intuiting how the stress will flow from the skin at the outer area of the blades to the spars in the inner section without anything but UNI running down the blades. I'm very interested in seeing the stress plots for the skins and the spars, as well as the interactions near the hub. I'm finishing up a couple of projects at work which should be done by the middle of January, so maybe I'll get a couple of hours over a few lunchtimes to try to set this model up. The "Mechanica" software we use has optimization capabilities, so we might actually be able to keep the safety factor up and make the structure somewhat more optimal. -- Marc J. Zeitlin Email: marcz@an.hp.com by InfoAve.Net (PMDF V5.1-8 #23426) with SMTP id <01IRLQNCVW1U9H8OPF@InfoAve.Net> for prop_fab@hpwarhw.an.hp.com; Thu, 25 Dec 1997 20:17:51 EST Date: Thu, 25 Dec 1997 20:17:50 -0500 (EST) Date-Warning: Date header was inserted by InfoAve.Net From: Nick Ugolini Subject: PROP-FAB: Spar Merry Christmas, Tomarrow I am planning to cut a dummy blade blank on my bandsaw (instead of using the chisel). I think I have come up with a easy way to cut the helix.... it worked on a piece of thick foam.... I will let you know how it works out. Now for a few thoughts. If we are going to use the spar (great idea as far as I am concerned) I wonder if it could be optimized to our advantage. Try and follow me on this one. A constant speed prop twists in flight. What if we used the spar in some manner to twist the prop to a greater extent.... If the spar was optimized (Marc you will have to help me on this) it should be able to cause the twist. Since the spar provides torsional strength, and torsional stress is dependent on rpm, and rpm is dependent on speed, the faster the blades turn the more stress is placed on the spar. If the majority of the spar was located at some point on the blade (fore or aft?) then wouldn't the blade be asymetericly loaded? Wouldn't the blade twist? Lets say the spar was mainly located near the LE on the top of the blade and the TE on the bottom of the blade? Or both top and bottom spars were place near the LE or TE? What effect would it have on the twist of the blade as stesses increase? With a safety factor of 5, the same amount of spar material the blade would have the same amount of strength, but..twist? More constant speed? Just a few Christmas thoughts for you to ponder. Nick Ugolini unick@mail.charleston.net Varieze N89RS LongEZ N29TM Cozy Mark IV #0264 From: Marc J. Zeitlin Subject: PROP-FAB: Spar (fwd) Date: Mon, 29 Dec 97 10:23:44 EST Nick Ugolini wrote; >A constant speed prop twists in flight. What if we used the spar in some >manner to twist the prop to a greater extent.... You've got the basic principle right, here, although it's usually the skin that resists the torsion (although not much, in the case of the straight UNI skin Nigel's got here). In this case, the twisting is resisted pretty much by the wood core, just like in any wooden prop. There are "almost CS props" that some MFG's advertise, and they work exactly as you theorize - as the RPM changes, the twist changes. At higher RPM, the pitch coarsens (static RPM is lower than at speed). >Lets say the spar was mainly located near the LE on the top of the blade and >the TE on the bottom of the blade? Or both top and bottom spars were place >near the LE or TE? What effect would it have on the twist of the blade as >stesses increase? Good questions, and not ones that have intuitively obvious answers (at least to me :-) ), since there's the aerodynamic loading as well as the inertial loading to consider. The FEA could determine this under different configurations. It might actually be necessary to modify the core somewhat to make it less torsionally stiff, and then add some tailored BID wraps to the skin to get the exact twisting characteristics that we'd want. >With a safety factor of 5, the same amount of spar material the blade would >have the same amount of strength, but..twist? More constant speed? Certainly possible - if the stress is almost completely caused be the inertial loads, then moving the spars around shouldn't change the safety factor appreciably. However, as previously mentioned, it's not the spars that are determining the torsional properties - it's the skin and the core. I'll try to keep this idea in mind when I get to the FEA. It's been done before on wood props with varying degrees of success, so it should certainly be possible. There was one company at OSH this year that was claiming to have a composite version of this kind of propeller for sale sometime in 1998, with exactly the properties you mention. They claim to have done a load of computer modeling to verify the characteristics. -- Marc J. Zeitlin Email: marcz@an.hp.com From: lschuler@cellular.uscc.com Date: Mon, 29 Dec 97 09:54:35 -0600 Subject: PROP-FAB: Spar Nick Ugolini wrote: >snip >If we are going to use the spar (great idea as far as I am concerned) I >wonder if it could be optimized to our advantage. Try and follow me on >this one. >snip >A constant speed prop twists in flight. What if we used the spar in some >manner to twist the prop to a greater extent.... Nick, I'm sure Nigel will jump in here after he gets back, but I think the main function of the 'spars' in this case is not as a typical spar or for torsion. I think tey are designed to keep the hub/blades together. That's why Nigel didn't run them parallel; he was trying to get more surface area in contact with the outer skin. Remember, these spars are also tapered to near zero depth at 40% of span. Most constatnt speed action (twisting) will be near the tips; not near the hub where the spars are. Larry From: lschuler@cellular.uscc.com Date: Mon, 29 Dec 97 10:21:05 -0600 Subject: PROP-FAB: Spar (fwd) Marc wrote: >snip >There was one company at OSH this year that was claiming to have a >composite version of this kind of propeller for sale sometime in 1998, >with exactly the properties you mention. They claim to have done a >load of computer modeling to verify the >characteristics. Went to their forum. Unfortuantely, they stated a number of times that the prop was not for use on a pusher.... Interesting. Sure looks sharp and sounds nice; except the price! Some competition for Hartzel CS, but not by much. Larry by InfoAve.Net (PMDF V5.1-8 #23426) with SMTP id <01IRR6NQ9ZPS9GFZ4D@InfoAve.Net> for prop_fab@hpwarhw.an.hp.com; Mon, 29 Dec 1997 17:50:13 EST Date: Mon, 29 Dec 1997 17:50:13 -0500 (EST) Date-Warning: Date header was inserted by InfoAve.Net From: Nick Ugolini Subject: Re: PROP-FAB: Spar (fwd) Marc J. Zeitlin >There was one company at OSH this year that was claiming to have a composite version of this kind of propeller for sale sometime in 1998, with exactly the properties you mention. They claim to have done a load of computer modeling to verify the characteristics.< I had a chance to visit the booth. The prop is basicly an injected molded prop, which very close in appearence to Klaus's (LightSpeed) prop. Individual blades attached at the hub. $3-4K per prop. Not being built for pushers... I trid cutting the helix on my band saw (test blade),saved lots of work.... It took me ~30 min to cut the bottom and top of the blade and shape the airfoil (I did not measure anything). I also designed a built in prop leveler and am working on a hand held duplicator design for the bottom blade. Purchased the wood today, and the rest of the material will be in soon..... Marc, if you think you will be able to do the modeling in ....let say the next month, I will hold off doing my layups until you are finished. It would be interesting to see if the design works. Nick Ugolini unick@mail.charleston.net Varieze N89RS LongEZ N29TM Cozy Mark IV #0264 Date: Tue, 30 Dec 1997 23:14:11 -0500 (EST) From: Nigel Field Subject: Re: PROP-FAB: Spar At 09:54 AM 12/29/97 -0600, Nick and Larry wrote: > >Nick Ugolini wrote: >>snip >>If we are going to use the spar (great idea as far as I am concerned) I >>wonder if it could be optimized to our advantage. Try and follow me on >>this one. >>snip >>A constant speed prop twists in flight. What if we used the spar in some >>manner to twist the prop to a greater extent.... > >Nick, I'm sure Nigel will jump in here after he gets back, but I think the >main function of the 'spars' in this case is not as a typical spar or for >torsion. I think tey are designed to keep the hub/blades together. That's >why Nigel didn't run them parallel; he was trying to get more surface area >in contact with the outer skin. Remember, these spars are also tapered to >near zero depth at 40% of span. > >Most constatnt speed action (twisting) will be near the tips; not near the >hub where the spars are. > >Larry Hi, Im back (finally). Interesting discussions re tailoring the constant speed effect. Having a little trouble conceptualizing how the spar structure would effect this though. As stated above and by Marc in another post, the spars only go out to 40% radius where the wood core is still pretty thick so there is little twisting happening inside the 40% radius. As Larry states most of it occurs in the outer 20 % or so where the work is done and the blades are thin. My understanding of how it works is that the blades need to twist a little in torsion proportional to the applied load, which is power, to provide some CS effect. Thats one of the reasons I used uni glass only span wise on the prototype. A bias ply adds little to longitudinal strength where the loads are highest but adds a lot of extra centrifugal load. However the plies on mine are alternated parallel to the LE then the TE so there is a little bit of bias increasing the torsional stiffness somewhat. So at high power as in take-off and climb where the blades produce high thrust (lift) the blades are allowed to twist to a lower angle near the tips. This is caused by the pitching moment of the airfoil which attempts to rotate the blade negative around an axis approximately at the .25 chord point just like a wing wants to do, only in wings we prevent this by employing a angular bias to the skins to give them stiffer torsional properties. But the biggest twisting effect is achieved by adding blade tip sweep back. This places the blade centre of lift, especially out at the tip, well behind the radial line resulting in a moment (lever) for the lift force to act on flexing the blade forward slightly and imparting a twist near the tip of the blade. The further back the sweep, the longer the moment and the more pronounced the twisting effect. The problem with this is that the blade mass out at the tip also acts on this same moment offset from the blade centre of mass creating very high bending stress on the blade in the rotational plane. Consider that a 4 lbd blade at 3300 RPM (overspeed) exerts 16.8K lbs of centrifugal force with the tip loads providing a substantial portion of this force (remember load is M*V^2/R). If you do the vectors, the bending loads in the rotational plane get pretty high. So if you sweep them way back the prop will really twist under load but it will also really want to break under the strain. Still these forces can be accomodated if there is sufficient structure provided, so a scimitar shaped prop can be made to work, after all there are plenty of them flying for some time now. I think Marc has the best approach to model it first in FEA and see what gives on the stresses before anyone flies one with a radical planform or structure. That should be a real challenge to model Marc, don't know how you would establish the modes and forces, true glass modulus etc., but it has been done so why not. I have Cadre FEA but don't think it powerfull enough for this job. Nick I would suggest that you build your first one on a conservative note, in other words keep it simple until you get a feel for how it works. You may be surprised as I was and find it works pretty darn good as it is. But everything can be improved upon and I think there are some exiting possibilities open to us here to get even better performance. Still I have often been painfully reminded that there is no free lunch in physics. I don't anticipate any dramatic improvements will be realized or someone way smarter than us would have done it already as props have been around a long time. The one at Osh that promises unheard of efficiency improvements is just theory and has yet to fly. Computer designed you say, well so was mine. As we all know a computer is just a fast adding machine controlled completely by human instructions. I will buy all of the beer for everyone next Osh if it achieves 1/2 of their efficiency improvement claims, but one never knows. Plan to start my next one very soon, so will document things this time and take some construction photos. Nick keep up the good work and please keep us updated on your construction progress. That band saw trick sounds great, may try that also. Nigel From: Marc J. Zeitlin Subject: Re: PROP-FAB: Spar (fwd) Date: Wed, 31 Dec 97 10:59:49 EST Nigel wrote; >My understanding of how it works is that the blades need to twist a little >in torsion proportional to the applied load, which is power, to provide some >CS effect. >So at high power as in take-off and climb where the blades produce high >thrust (lift) the blades are allowed to twist to a lower angle near the >tips. Correct me if I'm wrong, but I believe that the prop produces much more thrust at cruising speed than at TO/Climb (assuming a cruise prop, anyway). I think this is because of the lower RPM at TO/Climb, as well as the optimized Lift Coefficient for Cruise. I may very well be wrong here - I have not done the analysis, and this is another region where my intuition is weak. I wonder whether (given that we believe that the centrifugal loads are two orders of magnitude more important than the thrust loads) what actually causes the blade to twist is the aerodynamic loading under power (which is only a few hundred lbs at best) or the centrifugal loading as the RPM changes. As you state below, the "scimitar" tip can cause a great deal of twisting (at the cost of higher stress levels) but whether it's due to the offset thrust line (wouldn't the sweepback cause a _lessening_ of pitch with high loading - exactly the opposite of what you want?) or due to twisting caused by rotational speed changes - as the rotations speed increases, even just a few hundred RPM, the forces on the blade can change tremendously - isn't clear to me. Hopefully the FEA can shed some light on these issues (actually, a closed form solution could probably do just as well for a reasonable estimate). >But the biggest twisting effect is achieved by adding blade tip sweep back. >.......... The problem with this is that the blade mass >out at the tip also acts on this same moment offset from the blade centre of >mass creating very high bending stress on the blade in the rotational plane. >......... don't know how you would establish the modes and >forces, true glass modulus etc....... Not sure that I will. I'll make some simplifying assumptions about the wood and the glass stiffnesses and strengths, and then see what happens. If I can get order of magnitude results and general directions for optimization, I don't care what the exact numbers are. I just want to verify the _theory_ of what's happening here (where does the stress flow - in the skin or spar or both; how does the blade twist under load; what material could be removed to lighten the blades; how does changing the skin orientations affect the twisting of the blades, etc). We'll see :-). -- Marc J. Zeitlin Email: marcz@an.hp.com