|10-14-2012, 01:07 PM||#1|
Joined: Nov 2008
950 Adventure composite luggage system
Bits and pieces of my luggage system are here in various topics, but I decided to put in in one topic.
When I bought my 05.5 950 Adventure a few months ago, it came with the aluminium Zega cases and the OEM steel rack. I rode a few hunderd kilometers with this setup, and decided I did not like it at all. The LC8 is quite sensitive to weight on the rear, and especially far behind. Much more than an 1100GS at least. Also, the Zega setup is WIDE, which is a problem when lane splitting over the ever-narrowing Dutch freeways.
Since I built a luggage rack before for an LC4 which placed the luggage further to the center of the bike and lower which I liked very much, I decided to do so again. And since I am usually running the bike without side cases I decided to strip some weight at the same time. The OEM rack is a heavy piece of steel. Makes no sense: replace the stock mufflers with Akra's to save weight, and then add that weight again with the rack.
I investigated aluminium as the potential rack material, but I could not find a good deal on heat-threating the welded rack, and leaving it as-welded is tricky since the stronger aluminium alloys lose most of their good properties after welding.
I also tried to find some more exotic materials such as thin-walled PH17-7 tubing, without success.
And then, I decided to run a little test with aramid (Kevlar)/epoxy laminate. People say it cannot be done, but no argument was convincing enough for me. Which always makes for a good drive to do just that.
First test was impact resistance, one of the weakest properties of composites. Thus, I pulled some scrap aramid cloth from the box of 'leftovers', mixed up some epoxy, and wound the aramid together with some glass cloth (epoxy-glass is a far stronger bond than epoxy-aramid) over a piece of 3/4" PVC pipe. No effort was made to ensure a 'good' laminate, just hand layup.
Then the ultra-scientific test: drop a 20kg concrete block with it's edge on the pipe from about 1 meter (~3ft) height, and see what happens.
The glass outer layer fractures, the aramid stays fully intact. I repeated the test a few times every time curious frind dropped by, same result.
How does this compare to metal? Same test with a 28mm OD / 2mm (1.1" / .080") wall thickness 304 stainless pipe:
I also tried to overload the aramid tube by putting it verticaly on the concrete floor and load it. It bends a little, but it would not break.
Preliminary conclusion: aramid/epoxy will do. And I needed hands-on experience with composites to develop a 'feel' for the stuff, so I decided to copy the OEM rack in composite by laminating PVC tubing.
In hindsight this was not the smartest idea; a form more suited to composite construction would have saved a lot of time.
Bending up the loops using a wooden mold went OK using a little persuasion with the heat gun.
For modelling the mounting I used PVC sheet because it is such a nice material to shape using a little heat. PVC reacts very well to heating; a little heat and it gets a little more flexible, more heat and it gets more flexible. And it's a fairly continuous process which is easily controlled.
The mountings of the loops are designed to be more suited for composite contruction: large cross sections, thin walls.
Steel was used at the bolt locations to distribute the local stress a little.
PVC is heavy, so I lightened it a little
The formed PVC sheet was then 'thickened' with Depron; a very light (75kg/m3) but fairly strong kind of polystyrene foam, and a steel insert was placed at the bolt holes. Then, a cutting mold was made for the composite fabric.
Carbon fiber was used for the mounts, because it is stiffer, stronger and the mounts are not likely to receive a direct impact.
Now, when just laying up cloth and epoxy it does not result in the strongest possible laminate. A strong laminate needs as much fiber and as little epoxy as possible. The only way to archieve that is to compress the laminate and 'squeeze out' the excess epoxy. That is where the blue and white stuff is for. The blue stuff is perforated plastic foil which will pass the epoxy, the white stuff is absorber fabric which absorps excess epoxy like a tissue absorps spilled liquid.
For the compression I did not mess with a vacuum setup; I used shrink tape instead. It is the same stuff as shrink tube used for electrical connections, but then in tape-form and much thicker. When shrinking the tape, a lot of force is exerted.
The top mount received a lot of pre-tensioned aramid fiber, and carbon:
Next was laminating the loops. Since cloth has fibers in two directions running under/over each other, fiber density is limited anyway. So I decided to run a bundle of fibers along the loop's PVC pipe, finish it with a few layers of cloth, and wrap it tightly with aramid tow to compress it heavily.
This was a tedious job, but it worked like a charm. I calculated fiber density by laminate thickness and amount of fiber used, and came up with 60% fiber density. Those are autoclaved-formula1-part values
Since the surface was not smooth and optically perfect, I decided to paint the luggage rack black using a 2K polyurethane paint. I already have enough carbon fiber on the bike anyway
Weight of this rack: 593 grams / 21oz.
Number two was made a little different since this entire thing was also meant as a learing experience, and the first one really took a lot of time to construct which I wanted to reduce.
I also found out that strength is not the issue, but stiffness is. Contrary to popular belief carbon and aramid composites are far LESS stiff than metal with the same dimensions (as said before: I should have used a different design instead of a copy of the OEM rack). And stiffness is not improved by compacting the laminate more. The same is true for compression strength, also a weak point of composites. Compression strength is affected more by fiber density than stiffness, but not that much.
Thus, in number two I used more steel which saves a step or two in construction:
Much simpler steel reinforcements:
and just hybrid carbon/aramid fabric on the loops (pure carbon on the mounts).
Compression was done a little different this time also: I cut strips of thick polyethylene foil, and wound that tightly over the fabric. Excess epoxy could escape and the larger width made the winding process far less time consuming.
Technically, this laminate is not as good as the laminate on the first rack, but as I said before: (tensile) strength is not the problem, but stiffness and compression strength are.
The right-side rack weights 615 grams. Slightly more than the left-side rack.
OK. next are the cases.
|10-14-2012, 01:43 PM||#2|
Joined: Nov 2008
Since I use the cases only occasionally, I decided to use a fairly simple construction method which would not yield the lightest cases possible but lighter than regular aluminium.
construction method: flat composite sandwich panels glued into a box with aluminium profiles on the edges. Aramid on the outside to increase tearing resistance, glass on the inside to cut cost.
The case bottom panels were made of the following sandwich:
Holes were punched in the Depron and filled with glass strands+epoxy to connect the cloth layers together.
The side and top panels are basically half of this, thus aramid/carbon, aramid, Depron, glass, glass.
Weight of one of these 45x20cm panels: 150 grams.
More cutting of fabric. A lot of separate pieces must be cut for 2 cases. Takes the better part of an evening.
(this is not even halfway done)
Interface between the box and the lid. Standars flightcase-material:
Oh, i forgot to mention: I chose plywood instead of Depron for the core on the rear of the cases. That forms a much better support for the mounting system and load distribution than foam. Wood is an underrated, cheap and very nice construction material for these purposes.
Construction of the side/top panels.
I laminated them all, put them on top of each other with sheets of polyethylene and some breather in between, and compressed them by weighting them. There is not nearly enough pressure to make a 'good' laminate, but for case panels which are not loaded heavily it's not that important.
At least I found a good use for a case of Heineken beer.
The epoxy I use is very fluid and has a very long pot life of 210 minutes. Disadvantage: it won't cure enough in 24 hours and for full strength it needs a post-cure at elevated temps anyway. So after a night of roomtemp curing and with the missus away I accelerated the curing process:
Cutting excess fabric from the panels, cleaning, and prepare for painting:
Which did not work out like it should on all panels. Bummer. I hate sanding/filling/sanding/filling/etc.
Painting was done with a 1-part alkyd-urethane paint. About the most damage-resistane paint I can get in semi-gloss. I can get the far more durable 2-part urethane paint only in high gloss.
First few panels and aluminium profiles put together in a box-shape using lots of tape and thickened epoxy resin as the glue:
Few steps later:
How heavy are they?
2.7kg (6 lbs). More than I anticipated (I calculated the weight in advance to be 2.4kg), but still lighter than 30-liter aluminium cases.
Mounting system is simple and not yet 100% finished.
Carbon brackets for the bottom support:
Wooden 'hockey pucks' sawed out of a piece of plywood and reinforced by impregnating with epoxy, steel strips welded to M10 threaded rod, and wingnuts on the inside
Mounted on the bike:
There are some flaws here and there; I sanded into the aramid once (close to impossible to fix), the paint is only 99% smooth instead of 100%, etcetera. But the best thing is riding with them: the entire setup is around 9kg / 20lbs lighter than the OEM rack+Zega setup and the bike behaves a lot better with loaded cases than with the Zega's. Also nice: I sold the Zega cases only, built the rack and these cases, and still have money left in the pocket
|10-14-2012, 07:15 PM||#3|
Joined: Mar 2011
Location: Central Tejas
How many and what kind of miles do you have on this setup?
2007 KTM 950 Super Enduro A.K.A. Fireball the Austrian Honey Badger
2002 BMW 1150GS Adventure
veriest1 screwed with this post 10-15-2012 at 07:08 AM
|10-14-2012, 07:47 PM||#4|
Joined: Jun 2011
Location: Billings, Mt
Wow, Nice job! What type of latches are those? Look like southco cam latches I use on whitewater rafting aluminum dryboxes I weld up. I have been considering making some AL ones to replace my Gobi's.
Mttodd screwed with this post 10-14-2012 at 07:56 PM
|10-15-2012, 01:00 AM||#5|
Joined: Nov 2008
I probably might want to improve the mounting a little; currently it is a bit of a hack job since I needed the cases. Areas of improvement: more overlap on the circumference of the rack on the lower mounts and modify the top clamps to be able to hold the cases in the event that the bottom mounts break.
But I am sure that the cases and rack can withstand some serious abuse. The composite tend to bend a lot before breaking, which is a good thing in my opinion.
For the 'steel can be repaired in the middle of nowhere'-folks: it seems that whenever I break something the closest welder is a daytrip away. Composites can be repaired in the field using glass or cotton cloth and some quick-setting epoxy. And more than composites, so carrying some epoxy and glass cloth is useful anyway.
I am an expert in parking bikes horizontally and winter time with snow and ice is just around the corner, so I'll find out soon enough.
|10-15-2012, 02:19 AM||#6|
Joined: Oct 2009
Location: Hillsboro Oregon
Nice Job , kevlar is tough material
BTW , be very careful if you are spraying urethane automotive paint especially two part , the hardeners contain isocyanates that will cause permanent lung damage.
if not using a fresh air supplied respirator ( the kind with the hood) it could ruin your life , a cartridge filter respirator is not enough
04 KTM 950 S
04 KTM 525 EXC street plated
|10-16-2012, 01:52 AM||#8|
Joined: Nov 2008
Disadvantage: only available in high gloss and it is not that you can grab a can and start painting. Accurate weighting of the A and B components, mixing, leave it undisturbed for at least 30 minutes, thinning if required, and 45 minutes have elapsed before the first blob of paint can be applied.
|Thread Tools||Search this Thread|