Discussion Topic |
|
This thread has been locked |
mikeyschaefer
climber
Sport-o-land
|
|
|
Feb 18, 2017 - 09:57am PT
|
This is the most fascinating stuff I've read in a long. Keep it coming!
|
|
deuce4
climber
Hobart, Australia
|
|
|
Topic Author's Reply - Feb 18, 2017 - 10:51am PT
|
Actually, that RockWest data refers only to the carbon fibre fabric. This site looks like it has reasonable estimates of typical carbon fibre composite tubing properties: http://www.clearwatercomposites.com/resources/Properties-of-carbon-fiber
Using data from the Clearwater Composites site, let's add the high/low range of carbon fibre to our list:
A few things to note: People often just say "titanium" but the range of alloys is considerable with widely varying properties, as noted above in the Grade 5 vs Grade 9 titanium. Same with carbon fibre. The chart I used to get the carbon values has inconsistent data for other materials, i.e. a ultimate strength, rather than yield, for 4130, which is highly dependent on heat treatment, and lists less than the typical yield for 6061-T6, which is a very specific material and heat treatment. But all the metal material properties above are from the Matweb site, which is consistent (for 4130, I used the "normalised" specs, which is what Aircraft Spruce documents as the typical aricraft tubing treatment). So the carbon fibre specs above are to be taken with a grain of salt, but they seem reasonable (note also that carbon fibre strength decreases with increased Elastic Modulus). By the way, Aircraft Spruce is another great resource for learning about tubing and the common sizes available: https://www.aircraftspruce.com/categories/building_materials/bm/menus/me/index.html
|
|
matty
Trad climber
under the sea
|
|
|
Feb 18, 2017 - 10:58am PT
|
Thanks for sharing John!!
Really enjoy reading and learning, I'm a physicist and it's nice to get a peek into the more technical engineering stuff. Keep it coming!
|
|
nah000
climber
no/w/here
|
|
|
Feb 18, 2017 - 11:39am PT
|
yeah, don't be discouraged by the lack of responses...
just because there is less to comment on, when someone with actual knowledge takes the time to edumicate the rest of us plebs, doesn't mean there isn't an intrigued but silent audience...
looking forward to continued posts and thanks for all of the knowledge you've already shared...
|
|
Caveman
climber
Cumberland Plateau
|
|
|
Feb 18, 2017 - 11:47am PT
|
^^^^^^^ :)
|
|
stevep
Boulder climber
Salt Lake, UT
|
|
|
Feb 18, 2017 - 11:56am PT
|
I was thinking about that same joint issue with CF at the corners that you mentioned. Trying to connect bare tubes to metal corners wouldn't work very well.
But you could use industrial epoxy to bond metal ends on to the CF tubes and then joining those to metal corner pieces would seem doable.
I don't think straight-wall HM CF tubing is too expensive. Certainly cheaper than Ti. But might not be as cost effective as 7075-T6 Al.
|
|
nathanael
climber
CA
|
|
|
Feb 18, 2017 - 12:00pm PT
|
following closely, thanks
|
|
wayne burleson
climber
Amherst, MA
|
|
|
Feb 18, 2017 - 02:03pm PT
|
This is really good stuff John!
And better writing and simplified math than many engineering profs! :)
Looking forward to more.
-wayne (in Switzerland with my kids now)
|
|
looks easy from here
climber
Ben Lomond, CA
|
|
|
Feb 18, 2017 - 02:21pm PT
|
Another lurker checking in. I don't know enough to ask any reasonable questions, but I know enough to understand you and be interested.
|
|
deuce4
climber
Hobart, Australia
|
|
|
Topic Author's Reply - Feb 18, 2017 - 02:30pm PT
|
Wish I could "Like" all your posts, very encouraging. Thank you.
I am trying to create, for the benefit of the community (including other manufacturers!), but also for myself, a clear and concise way to think about these engineering aspects. As an aside, I look at my old notes from the A5 days, and though they served me well to optimise the design of the time, it's hard even for me to make heads or tails of them these days. For example, below is a reference chart I used for years to consider other tubing sizes, as well as joints and other aspects of portaledge tubing design (i.e. double butting):
Or this when I was looking into Titanium:
But I am hoping to make it more clear now. Just to note, a lot of my recent work and conclusions came from playing with FEA (Finite Element Analysis) tools such as the quick analysis I did in the stress/displacement pics below, but analysing using FEA is more an interactive process and actually challenging to document in a clear way (see previous FEA video for example, still probably a bit too much "engineering" stuff).
But I digress (again!). And I am giving away more than I wanted to this point--was hoping for more of a surprise at the solution of eliminating spreader bars, but which I will get to soon.
to be continued...
EDIT: here's a link to an amazing tool--Autocad Fusion. I used a lot of different FEA tool in the past 15 years of my engineering, starting with Patran/Nastran (not user friendly!) when I was studying for my Masters of Engineering at UNSW in 2003, but Autocad Inventor is really, really good for relatively easy analysis (for design, I prefer Autocad Inventor, but Fusion is ok for solid modelling, too).
http://www.autodesk.com/products/fusion-360/overview
|
|
Rhodo-Router
Gym climber
sawatch choss
|
|
|
Feb 18, 2017 - 02:35pm PT
|
I met a guy who makes bamboo bicycle frames with CF corners.
|
|
duncan
climber
London, UK
|
|
|
Feb 18, 2017 - 03:03pm PT
|
Have I lost everyone? Seems like the thread isn't getting much response lately.
Watching and thinking! To increase rigidity you've changed some combination of the frame material, or the frame cross-section, or the frame design in some way.
You've implicitly ruled out Ti as a material which leaves carbon fiber or larger diameter aluminum tube. As I was writing this you've discussed carbon fiber composite, which was my original guess, but that seems not on the table for now. Larger diameter, thinner walled, aluminum tubing would be stiffer for the same weight, but be easier to dent. Would this work?
Regarding frame cross-section, from the picture and all your sketches it looks like you're still using round section tubes, so not something like I-beam. I guess I-beams would be better at resisting bowing if it was a simple beam application but it's not. Oval cross-section tubes, angled down and in, so the direction of pull of the bed material is along the long axis of the oval? These would be stiffer than round section tube but harder to implement, certainly if you continue to use corner blocks.
The bed shape is still rectangular, so not something fundamentally different like an oval (you have a pesky rock wall to deal with, most of the time). I wondered if you could somehow preferentially load the short side of the rectangle with the bed material, a bit like a hammock, but couldn't think of how you could do this and it wouldn't work with a bunch of people all on the same ledge. This suggests you've changed the frame design in some other way. Could the long sides be pre-curved - bowed outwards subtly - so that loading the bed material would bend it back to a rectangle - or would this make assembling too difficult?
Interested to hear about the solution.
|
|
deuce4
climber
Hobart, Australia
|
|
|
Topic Author's Reply - Feb 18, 2017 - 03:08pm PT
|
Nice input Duncan, and cool ideas.
The solution is actually much simpler than you'd imagine! But it works, and keeps the ledge light.
When I eventually reveal the solution to eliminating the spreader bar (and an easy way for companies using the block-corner design to revamp production without having to redesign the fabric parts--and save on cost, too), it will likely seem obvious. But I guess not so obvious that for the past 20 years the two major producers of portaledges have opted for a spreader bar solution to flex ;).
By the way, this structural aspect is only one of the four main innovations of my new design, the D4 Portaledge.
back soon.
|
|
BLUEBLOCR
Social climber
joshua tree
|
|
|
Feb 18, 2017 - 04:05pm PT
|
You did crew John. Those CF oars are strong and light as shite. The oars we had at UCSB were 12'-9" X +\-2" dia. Granted a couple feet of that are the blade and handle. Made from balsa the handle's are lathed to fit inside the CF tube about 2". Maybe for a ledge, instead of the tubes goin inside the corner's. The corner goes inside the tubes?
i like your new logo, that's bitchn bob😎
|
|
Moof
Big Wall climber
Orygun
|
|
|
Feb 18, 2017 - 06:30pm PT
|
Bump for the best non-Trump thing on here.
A few years back a one many shop called Burley Equipment made an A5 knock off with a 2:1 on the strap (also made a burley "Love Swing"). They disappeared after less than a year. Seems like something that needs to become standard. Dialing in the ledge flatness on a double is a real pain with a fatso like me sharing it.
For my CF ledge I double butted the ends with some 0.058" 1" OD for fear of the ends shattering under rough use and used stuctural epoxy to bond them (expensive stuff). At the side joints I put external as well. Added more weight than I should have, but without much of an ME background or any mechanical FEA available, what are you going to do?
BTW, is that Ansys in your screenshots? Any recommendations on hobby budget FEA tools that are worth the hassle and money? I use FEM electromagnetic tools in my day job (including Ansys HFSS), so I have a rough idea how ugly the learning curve can be.
|
|
edavidso
Trad climber
Oakland, CA
|
|
|
Feb 18, 2017 - 09:55pm PT
|
As an mechanical engineer myself, I'm really enjoying this thread. Also interested to see where it is going. A couple of comments on the technical discussion:
Your numbers for the carbon fiber tubes are not that far off. Carbon fiber, however, has a few differences from metals that should be taken into account. One is that it is an orthotropic material, meaning its material properties are different in three orthogonal directions. Metals are considered isotropic (material properties consistent regardless of direction). For a tube in bending, this means the bending stiffness will be a little lower than the modulus in the longitudinal direction indicates. Another relevant difference is that unlike metals, which generally have the same strength in tension and compression, the compressive strength of composites is less than the tensile strength. A tube in bending will be in compression on one side and tension on the other so the design should be based on the compressive strength. For HS carbon fiber (low modulus), this will be closer to 200 ksi. For HM carbon fiber, it will be around 160 ksi. In either case, the modulus and strength will depend on the exact layup and fiber used as you've previously mentioned.
To simplify the technical discussion, you could use thin wall approximations for the second moment of area (area moment of inertia). For a thin wall tube, I = PI*R^3*t. This is a little more intuitive and will actually be pretty close for the calcs for these tubes.
I like to think of stiffness in terms of load per deflection, just like a spring. Many are familiar with F = k*x for a spring, where F is the load, k is the stiffness and x is the deflection. You can treat a beam in bending in the same manner. Stiffness, k, then is proportional to E*d^3*t/l^3, where E is modulus, d is diameter, t is tube thickness and l is tube length. This shows that stiffness for a tube in bending is a combination of a material property (E) and geometrical properties (d, t, and l) and that d and l are the primary contributors as they are cubed.
More important for this discussion might be bending stiffness per tube weight. Using thin wall approximations for "I" and weight, you get that k/W is proportional to E*d^2/(l^4*rho), where rho is the material density. Again, this is a combination of material and geometrical properties, for material the important parameter is (E/rho) and for geometry it boils down to d^2/l^4. Intuitively this should make sense - to increase stiffness per weight you want to increase modulus, decrease density, increase tube diameter, and decrease tube length.
Sorry, that ran a little longer than I wanted but hopefully it made some of the engineering concepts a little easier to understand for some.
|
|
BLUEBLOCR
Social climber
joshua tree
|
|
|
Feb 18, 2017 - 10:41pm PT
|
^^^that was cool, thanks:)
|
|
deuce4
climber
Hobart, Australia
|
|
|
Topic Author's Reply - Feb 19, 2017 - 02:27pm PT
|
Making some progress on presentation. Here is a chart showing relative strengths of tubing. Note that this does not take into account the length of the frame, which is important specification for the bending stresses, I will try to incorporate that later.
The "Strength Index" incorporates the engineering aspects, and is useful for comparison.
The above chart in bar graph form:
I consider the first tubing, 1.125"OD, 0.058" 6061-T6 aluminium the baseline--it is certainly the material that most portaledges (probably going on 10,000 ledges by now) have been made from, and has been proven to work well without spreader bar for a 43" x 75" ledge.
But remember that we're not just looking at strength--we need to consider flex as well. Coming soon...
By the way, I am just now experimenting with carbon fibre again (those sizes listed above)--need to glue or rivet some aluminium ferules to the ends of the carbon fibre--any suggestions?
Also, here's edavidso's boiled down formula from latex editor:
Intuitively this should make sense - to increase stiffness per weight you want to increase modulus, decrease density, increase tube diameter, and decrease tube length.
Good thing to keep in mind for next part of discussion...
Edit: not sure if E, the elastic modulus (aka Young's Modulus) has been properly explained--will try to include...
Edit2: revised A5 Great Trango Titanium ledge--measured it, it is actually 0.035" wall (the one for Catherine Destivelle was the stronger, thinner wall 6-4 material)
|
|
Moof
Big Wall climber
Orygun
|
|
|
Feb 19, 2017 - 03:45pm PT
|
I used Hysol 9340 structural adhesive for my CF ledge. Seems to be burly stuff.
A CF tip: If re-inforcing the corners internally like I did (paranoia...) pre-cut an oversized hole in the aluminum, then use a diamond tip hole saw for the bungie hole. Diamond drill works great on CF, badly on aluminum. CF is terrible on non-diamond tips.
Oh yeah, measure twice, then measure again, then mockup, then cut your CF. CF is REALLY pricey to eff up on.
I am all for smaller ledges too. The hard part for me was wanting to just buy a fly. I got a couple A5 Cabana ones, and a simple ACE one for cheap, so it is far easier to make the ledge match the cheap flys than to deal with that much more seamstressing.
I am diddling around with tensioning ideas. I want to try using about a dozen or more small d-rings with some 1/2" webbing (smaller?) to really be able to get some mechanical advantage when bed tensioning and to spread out the stress. The original A5 patterns your posted had the reinforcing triangles aligned with the pattern of the bed (not on the bias). This creates quite a stress riser at the interface between the bed and the triangle. Cutting these on the bias (or better, use diamond patches) would better use the patches to spread the tension over a wider area of the bed making stories of knees poking through beds less likely.
On the bed sides I found that edge binding sucks, and for 2 reasons. First it is hard for small one time operations to get it right (pre-iron the tape in half!). Second, if you edge bind and then just fold/sew to make the passage for the frame half tension is held by the very end weave of the bed, yikes! So I use an attachment that does 1/2" single edge bind (folds over and sews, no tape), then fold that over and sew. It makes it much harder for the stitches to pull through under load. You also then only have to tape bind the corners and cutouts, which are much shorter and easier for hobby grade fooling around.
Wayback link to old discussions on bigwalls.net:
http://www.bigwalls.net/forumStatic/index-topic=217.msg1570.php.html#msg1570
|
|
|
SuperTopo on the Web
|
Let us know!