Why Did Bernard Kerr's Prototype Frame Break at Crankworx Rotorua?
A catastrophic frame failure in the middle of a race run is a nightmare scenario for any professional athlete, and for the company that manufactured the frame. Unfortunately, that's exactly what happened to Bernard Kerr and his prototype Pivot during the Crankworx Rotorua DH race this past weekend. Towards the bottom of his run he pulled up for a triple and came up short, with the front end of his bike separating on impact, as the above video demonstrates. Thankfully, Bernard wasn't seriously injured, but it's still the last thing anyone wants to see happen.
Chris Cocalis, Pivot's founder and CEO, provided the following insight into what likely caused the frame to break:
 | After a preliminary investigation into the failure of the prototype frame, we have identified that there was no failure of the carbon tubes and the epoxy used to bond the frame was still intact on the carbon tube side. Unfortunately, virtually no adhesive remained bonded to the aluminum. We use a treatment on all the aluminum parts to protect from galvanic corrosion between the carbon and the aluminum. We believe there was a failure of this treatment to properly etch/adhere to the aluminum. We have additional frames that were made in the same batch, which we will be putting through full destructive testing as well as getting all the frames from the team back that were produced at the same time. Once we have the opportunity to test these prototype frames, we will take the necessary steps to ensure that the frames are as safe as possible, and ready for the full impact of another World Cup season. We never want to see a frame failure and certainly not a failure of this magnitude. We are incredibly thankful that Bernard was not seriously injured and are taking steps to make sure that this is not an issue in the future. |
One of the prototype frames being assembled at Pivot's Phoenix, Arizona, headquarters.
Racers often serve as test pilots, putting hard miles in on bikes and equipment before they're launched to the public. Lab testing can only go so far, which is why real world testing is still an integral part of the development process. In this case, Pivot has been developing the new DH bike out in the open, allowing the general public to trace its progress over the course of last year's World Cup DH season.
The frame that broke had been raced at Hardline and seen numerous practice runs put on it in the lead-up to the Crankworx DH finals, so it hadn't exactly had an easy start to its life. Still, an instance like this does illustrate just how quickly things can go wrong. It is good to see that Pivot are already working to correct the issue, and we look forward to seeing Bernard and the rest of the Pivot DH team on track when the World Cup season kicks off at Fort William.
BK sending a whopper at Crankworx Rotorua before things took a turn for the worse.
Author Info:
Must Read This Week
Manufacturing is what (hopefully) their Asian suppliers are good at.
Completely right.
This was a production fault.
The first flaw is the straight-insertion of the top tube AND the down tube into the lug, at the same draw angle. As such, if the bond isn’t perfect and the bondline fractures, the tubes are not mechanically constrained by different draw angles...both tubes pull out of the lug(s) in the exact same draw direction — that is a serious design / engineering flaw in my professional (composites engineering) opinion. If designed differently, with different draw angles for each tube / lug, both tubes likely would not experience the same stress vectors at the bondine (due to their varying orientation) and thus both bondlines likely wouldn’t fail at the same time. And both tubes would not (and could not) pull out of the lug(s) at the same time, in the same direction. With different draw angles / bondline orientation / stress vectors, one bondline would likely fail before the other, allowing detection of the bond failure. Likewise, the different draw angles (if the tube are stiff & strong enough) wouldn’t allow both tubes to pull out at the same time, in the same direction. Atherton got that right. Pivot got that wrong. It was a major oversight and a fundamental engineering flaw, in my opinion. That Pivot frame even has a separate individual top tube, and a separate individual down tube — so they didn’t need to utilize the same draw angle for both the top tube & down tube. (But the same draw angle would be required if the rest of front triangle was a single-piece composite molding and the top and down tubes weren’t separate pieces).
Also, if the bondline wasn’t perfect, bonding failures can occur from bonding adhesive microcracking, adhesive macro-cracking, and incomplete adhesive coverage. Galvanic corrosion, and other corrosion, can be an issue with aluminum lugs and carbon fiber (unless the carbon fiber is insulated with fiberglass or similar non-conductive material). Aerospace companies often use fiberglass layers to prevent galvanic corrosion, making sure to not sand them off, nor sand too deeply into the fiberglass, where it bonds to aluminum. Likewise, at a minimum, they’ll typically hard anodize the aluminum, or use another bonding-optimized mil-spec surface preparation or treatment and/or anodizing process. Often, aerospace companies apply a military-spec type of cured epoxy primer onto the aluminum, to which the subsequent prepreg composite, or bonding adhesive, can tenaciously bond (and bond without corrosion due to the protective anti-corrosion epoxy coating). It appears Pivot excluded some or all of these critical (and fairly simple) steps. I also don’t see any fiberglass on the cured tubes, in the bonding area, in the prototype manufacturing pictures that Pivot released for publication on Pinkbike — it appears to be solely a woven carbon fiber surface layer, with no galvanic insulation layer, in the area bonded to the lug.
Additionally, it appears the bonding process was flawed — I didn’t seem to see much adhesive on the composite nor aluminum portions of the failed bond (in Kerr’s broken frame pictures). There’s a chance they didn’t properly remove mold release on the cured composite tubes. Or that they didn’t properly prepare the composite & aluminum surfaces (via abrasion and solvent cleaning), or that insufficient bonding adhesive was used, or that the adhesive scraped out of place during the assembly process...thus resulting in insufficient adhesive in the area to be bonded. The bonding process also appears suspect, from what I’ve seen in various pictures of the frame & manufacturing process. Trek (OCLV) and others co-mold several 0.002” tall “pips” around the circumference and length of their “male” bonded tubes...those pips center the tube into the female lug (or tube) and ensure an optimal 0.002” thick bondline, and it ensures that the adhesive doesn’t get scraped out of place as the tube is inserted. I doubt Pivot did anything like those 2-thou pips, based on several of my observations. It’s also usually a good idea to “inject” a bit of structural epoxy bonding adhesive into a port at the bonded joint, after initial assembly with adhesive, to ensure the entire bondline is fully covered, saturated, and void-free. I don’t see that sort of adhesive-injection port feature on their frame (from my cursory review), but I think Specialized might do that on their prototype DH frames. I think Atherton does the same. It’s common in aerospace, F1, and even in the production of certain composite sports cars. I actually import special 3M structural adhesive from the UK (it’s what I used on McLaren’s carbon fiber “tubs” and Bentley Motors’ prototype carbon fiber doors for one of their Super Sports technology demonstrator cars) which is better than what’s available in the US (likely because it’s more toxic =) — it’s strong, and has superior mechanical & dynamic performance numbers, and is highly toughened (as well as being highly crosslinked for extra strength, if you post-cure it a bit around 250F to 350F after achieving initial bond strength at room temp, overnight. Engineers, prototype manufacturing technicians, and production manufacturing technicians, cannot mess around with bond failures — it has to be done right, and the materials and manufacturing processes have to enable bonding perfection. Every time.
Also, Pivot appears to use cast-silicone custom bladders for their prototypes — and silicone / silicone oil can readily leach into the prepreg (unless significant protective countermeasures are taken). The silicone contamination can greatly adversely affecting bonding / adhesion. Aerospace companies ban everything silicone in their composites manufacturing facilities, for this reason. Even those yellow silicone “Livestrong” Lance Armstrong bracelets are strictly banned. Pivot’s custom composite tubes could have been contaminated with silicone that could affect their adhesive-bond performance — and their manufacturing pictures show the silicone bladder directly in contact with the uncured & cured carbon fiber prepreg, directly adjacent to the end of the tube (where the composite tube is subsequently bonded). Polyurethane, latex, molded nylon, or nylon film over polystyrene foam rigidized-bladder mandrels, are good alternatives for prototypes (and sometimes production).
Lastly, the kink / bend in the carbon fiber downtube, just behind the lug, will put both sides of the downtube in compression right at the kink / downtube bend...and carbon fiber fails in compression. Although Kerr’s bike didn’t fail there, it’s a suboptimal compression-loading scenario which will result in pronounced compressive forces in the laminate on both the top and bottom sides of the downtube at the bend (in a jump-case loading scenario like Kerr experienced). Frames have to be engineered, in every way, for those worst-case cases...and engineering a frame with a tube geometrical configuration that results in substantial compressive forces on both sides of a composite tube is just questionable engineering, in my opinion. The goal should be to limit inducing those sort of compressive loads and mitigate the risk of composite compressive failure.
Also, Atherton uses titanium lugs while Pivot used aluminum lugs. I'm no expert and just a guess, but could the titanium being more resistant to corrosion than alumunium be a factor?
Do you know what adhesive atherton bikes or Spezialized uses for their prototypes ?
youtu.be/j2UqzYQ_Qjc?si=_MQ_ED50CCE0_rTF
I think a lot of bike may of cracked or broken with the amount of forces that went through that frame and the angle of the landing, these things happen look at GT brand new DH bike it snapped clean in half in front of the world but people like Danny Hart are riding it without the same issues and Greg's V10 a few years ago I know it hit a tree but look how many other bikes (even the same models) have had similar crashes and nothing has happened. It was a freak accident that happned on the world stage the amount of bikes that break while testing and not being seen would probably shock everyone but you don't get to see or hear about it
Honestly I think large frame manufacturers are having troubles sourcing the right talent.
A failure of this magnitude could have caused the death of an athlete, then I would like to see their sorry faces.
I _do_ agree that in racing there is a risk of prototypes breaking, but fundamental process problems like this case are _not_ what you want as a point of failure. Carbon layups and bonding are not garage-sport technologies and the companies involved in them need to have a certain amount of resources available to do things right. How Pivot analyzes, improves and moves on will say a lot about their engineering/manufacturing credentials.
Last year there was at least one full head tube separation on a production frame.
But I’ll tell ya, despite my prototype Pivot DH bike engineering thrashing, I’ve utterly thrashed my own Pivot Mach 4 with zero failures in 7+ years — it’s a 28.5 pound (13.0kg), with pedals, aluminum 26” Pivot Mach 4 cross country bike that I ride like a DH and/or slalom bike (the Avalanche cartridge in the 120mm fork really helps =). I still can’t believe how well that bike rides...climbs well and goes downhill even better. Then again, Pivot’s M4X (4X / slalom frame) was just a slightly modified Mach 4. Some pics & vids below!
Haleakala Schleyble-Top:
www.pinkbike.com/photo/26370511
Greenlake wheelie fake?
www.pinkbike.com/photo/26370508
Maui Waui manual-lander:
www.pinkbike.com/photo/26370509
My best Myles Rockwell pose (no Maui Waui involved tho):
www.pinkbike.com/photo/26370510
Never lift when you drift:
www.pinkbike.com/photo/26370512
Trying to destroy my Mach 4:
www.pinkbike.com/video/524822
You steer a bike with the rear wheel:
www.pinkbike.com/video/524791
I have designs (and most importantly the process control developed and defined) in space applications and commercial planes that rely on surface preparation of aluminum for adhesion. We normally use plasma etching, a primer and a curing oven before bonding). The loads and the lifetime requirements are pretty challenging. I guess that on a MTB those challenges are even more difficult ... plus the consequences.
How could we trust anything they "innovate" with future designs, real question.
www.pinkbike.com/photo/26370709/?s6
The frame failure is similar to what happens when too much primer is used, surface contamination (silicone), or lack of basic surface prep (one could argue all of these are examples of).
Pivot said they believe the failure was the material used to limit corrosion. Amazingly it sounds like they didnt do lap shear testing of various material, thicknesses, and surface prep. Or if they did, it sounds like they need to update process and procedures for the technicians.
Ultimately, this is just an example of new engineering, failures happen and we learn. Glad Kerr is not hurt, want to see this bad ass get the #1win !!!
I noticed between the Atherton and Pivot, their cable ports are in different locations. The Pivot looks like it's lost 30-40% bonding area on the down tube and Lug.
What are you thoughts on that? Seem like a complete oversight.
Perhaps they went with the sketchy parallel straight entry top and downtube to allow for front center/ reach geometry changes without changing the headtube CNC assembly dimensions?
With an Atherton I imagine they have a headtube lug CAD design for each of their geometry options, because the angles between top tube and downtube change with different reaches.
This way they can change reach without having to make a new headtube shape each time.
Maybe that's the point of the questionable bent down tube as well? The bend allows them to just cut the down tube to the desired front center/ reach?
Good to read some good old german engineering thoughts on this one.
As far as the "prototypes break" that's very true. But in 2024 the fatigue/failure points of carbon and aluminum are pretty well known. In years past it was "hey the frame failed but we learned something and are adjusting our layup." Those failures make the finished product stronger or even carry over into early generations of the finished product (v1 Yeti Sb6). This failure isn't like that. It's a defect that will not help or hurt the finished product. It's embarrassing for Pivot and could have been catastrophic for Kerr.
At the same time people saying well Yeti's prototype didn't break. Yeti's prototype is way more finished than this one. Living in Arizona and seeing prototypes around this is a very early version.
I think both sides of the argument are looking way too much into this. It's a big deal for Pivot and Kerr and hopefully he is okay. But it doesn't really have an affect on the product we will be offered as the consumer.
But not all composite-tubed lugged bikes broke — Colnago’s C40, with its carbon lugs and carbon tubes, is one of the most well-liked road bikes of all time (including by TdF champs such as Bradley Wiggins — he and I battled for a 63cm NOS 1994 Colnago C40 w/full Dura Ace on eBay...and yeah, he really wanted it, and he won it, haha). Still hoping to get a C40 at some point (especially as my 64cm full-6/4-Ti 1999 Colnago Oval Master Titanio former team bike is my favorite handing, and favorite ride-quality, road bike of all time...having not ridden a C40 yet =)
Luckily Kerr is fine, but can you imagine the lawsuits coming Pivot's way if these bikes made it out into the wild and someone got kebab'd by a toptube?
And stop with your anti galvanic corrosion BS. Its titanium lugs bonded to carbon fiber tubes. Dynamic assembly break, EVERY design has a cyclical load. Meaning, you shake that shake weight enough times, it will break.
Please actually get a design and manufacturing job before you sprout non sense.
—————
Oh, we did a lot of the engineering, R&D, and prototyping work in Seattle, and some at our facility in Andover UK, but I went to the McLaren Technology Centre in Woking a lot — all from 2005-2010. Then I did some additional McLaren Applied Technologies projects for several years.
As for the carbon tub, we backed out of producing the tub/chassis for McLaren and Carbotech in Austria took over when we didn’t want to take on the financial risk associated with McLaren’s quite-optimistic sales projections. Apparently Carbotech was almost sunk because of that issue (specifically, the lower cash flow from lower volumes in the first years made it difficult to recoup their investment and difficult to cover high overheard costs associated with the program). That’s one reason why McLaren moved production in-house to their new(ish) custom-built Sheffield facility a few years ago. The “patented” 4-hour “single piece” manufacturing was my contribution =). It’s actually a lot more than 4 hours — it’s just a 4-hour mold utilization cycle...with dozens of hours of other labor prior to, and after, that 4-hour mold cycle time.
The crazy thing is, I was supposed to be a summer manufacturing intern on the McLaren SLR in 2005 — it took about 9 months to arrange the internship with McLaren, being from the US. But then I broke my back snowboarding and had a spinal cord injury, a few months before the internship was due to start. I called McLaren’s HR manager from the hospital and she told me I was SOL — the internship wouldn’t happen due to my required recovery time. I was crushed. I didn’t know what to do if think, nor did I know if I was going to be able to walk, so I simply prayed for an opportunity to work with McLaren in the future. Amazingly, that put me at peace about it.
A few days later my friend Nichole visited me in the hospital...super cool and smart gal who loved cars and had a rad VR6 GTI. We got to be better friends during my recovery, and one day a few months later she told me her dad does aerospace composites, kind of like the automotive composites I was keen on, and that I should meet him. When I met him, he offered me a tour of their Seattle aerospace composites facility. A few weeks later the tour day was approaching and his engineering manager called me and told me to bring my resumé. I was stunned — my medical insurance had changed and I was going to no longer be receiving medical coverage in a few months, and at the time I wouldn’t be able to get new independent medical insurance due to my “preexisting” spinal cord injury (a rule which changed with Obamacare). But the tour went well, they liked my resumé, and they hired me a few weeks later — and I got full medical coverage, despite my “preexisting” (and expensive) SCI.
But 8 weeks after starting at that aerospace company, the sales manager called me to his office to discuss new market opportunities like nanocellular foams + composite skins, and single-crystal sapphire bulletproof armor windows with bulletproof composite doors for military vehicles. But when I was waiting outside his office, I heard him on a phone call discussing “McLaren.” I was mystified. When the call ended, he called me into his office and I asked about “McLaren.” He said “Oh no, you don’t know these people, they’re in England.” And I said, “No, I do know them, I was supposed to be working there as a college intern right now!” He was just as surprised as I was. Turns out McLaren had come knocking on our door (well, through the phone), as our company’s UK site made the carbon fiber fabrics and flat-pack CNC-knife-cut carbon fiber preform kits for the McLaren SLR program. McLaren wanted us to bid on their new car program...the program for engineering and prototyping and manufacturing the carbon tub for their next vehicle (just like is used on all McLarens today!). It was a 5-year R&D + Prototyping program, and then a $250,000,000 multi-year manufacturing contract =o ! We were all amazed by the situation — they had hired me to help them get into new markets, but they didn’t expect a perfect fit like this!
Two weeks after that, I was at McLaren Technology Centre bear London giving a sales / technical / business proposal presentation with our team, to the senior engineering & business leadership of McLaren — and after talking all lunch with their otherwise-reportedly-mute engineering director (where we accidentally invented McLaren’s hydraulically-controlled suspension-roll-control-system based on my similar Formula SAE system), we ended up being awarded the R&D contract! I couldn’t believe it.
The next day, we went to the McLaren SLR production facility in Portsmouth. While there, the manufacturing engineer who was supposed to be my internship mentor told me that the HR manager, Lynne, was in the office. We both realized we had to go talk to her, as just a few months earlier, she had to end my internship...all while I laid in the hospital, unable to walk. I walked in, Graham was smiling, and I introduced myself and Lynne had a look on her face like: “I told you no months ago, and I’m not changing my mind, I don’t care how far you flew to talk to me.” But then I told her that McLaren came knocking on our door in Seattle and that I was going to be the lead R&D engineer on manufacturing their next new McLaren carbon fiber vehicle structure. She couldn’t get her head around it. It made no sense to hear. Graham was just laughing at the stupefied, perplexed, almost-alarmed look on her face. She asked how this could have possibly happened. I told her I didn’t know, other than that I had prayed for an opportunity to work with McLaren, moments after I had last talked to her, when the internship fell through. None of us had ever experienced anything like that before, so we all just talked through it a bit...it was a fun conversation =). But she was just amazed that I was supposed to be a lowly production intern, making SLR’s all day for minimal pay, and now I was in charge of development their next carbon fiber car structure that would replace their SLR and replace their Portsmouth facility. It still doesn’t make sense to me today, aside from a direct answer to prayer in a crazy semi-miraculous way...including learning to walk again just months before! But yeah, I’m still waiting on that McLaren 720S in British Racing Green prayer =P
It's a shame that the McLaren brand has been so mismanaged. Build quality issues aside, the cars are on a different planet to anything from Maranello etc.
The Atherton’s have it dialled.
The aluminium lugs are 1mm thick at best and have literally torn apart.
FAR too thin to be safe in my opinion as a fabricator. (admittedly im not an engineer so take that as you will)
It looks as though the lugs tore apart as a result of the bottom tube coming out of the lug and putting all the downward pressure onto the top tube.
You can inspect welds after testing for your QC, with bonding you basically do statistical spot checks (I assume on bikes also the welds checked are only a small percentage).
Does Pivot also work with double shear lap joints like Atherton/Robotbike uses? These lugs look machined so I suppose it would be too hard to do. The primary reason for using that is to avoid delamination at the free edge (where the tube got trimmed) but it may also give some redundancy as the tube is bonded both on the in- as well as the outside. They have a small hole on the outside for air and later the excess glue to evacuate so at least they know the glue is everywhere. So yeah, that's the check they can do. But I don't think they can actually check whether the glue has properly bonded to the metal. That said, I know that they have the Fokker Bond Tester (a resonance test) for checking glued connection in aviation but that's for body panels. Not sure whether it could be used for relatively compact machined parts.
Either way, failures like these are scary and I'm happy that Bernard came off relatively ok. Back in the days we built a car for a race in Australia where we built the suspension linkage out of pulltruded carbon tubes bonded to aluminum connectors glued to the inside (which were then connected to the bearings). During a test drive on Arnhem Highway (in Arnhem land, just east of Darwin) one of those glued connections failed causing a crash that could have ended much, much worse.
So yeah, carbon production has come a long way but it is still far from perfect. Remember though that head tube failures on aluminum frames used to be quite common whereas I haven't heard of (m)any in the past few years. Frames like these will get there someday as well.
Sure they have it nailed.
Etching cleans and roughens the surface, this essentially gives the material something to stick to, everything bonds better with some profile, some porousness and aluminum is not a very porous metal.
From experience in the coating industry aluminum and galvanized are hard to coat and require a ton of prep and specific products to provide a long life.
When they say etching failed, you can use a mechanical etch like lasers, or a chemical etch like muriatic acid. either the mechanical or chemical etch was insufficient to provide a profile to bond to.
Many products struggle with bonding issues, personally the raceface next sl cranks came to mind as the insets suffered multiple bonding issues and ultimately a recall on production products.
Titanium: By looking at the standard electrochemical potential of titanium, it seems that this metal is an active metal. However because of the formation of a dense stable and protective oxide layer, titanium is placed among the noble materials and just below graphite or carbon in the galvanic series table. (For a primer, see the article An Introduction to the Galvanic Series: Galvanic Compatibility and Corrosion). Therefore, there is no significant gap between titanium and carbon-fiber-reinforced composite to create galvanic corrosion. This means that commercially pure titanium and its alloys are completely resistant to galvanic corrosion when they are coupled with carbon composites.
www.corrosionpedia.com/galvanic-corrosion-of-metals-connected-to-carbon-fiber-reinforced-polymers/2/1556
Great that Pivot is being so open, this is what it is all about.
Obviously didn’t want this to happen to anyone, but it is a perfect example of what can happen sand as Chris quite rightly points out Bernard’s bike had already been ridden super hard at hardline!
Cheers
Basic synopsis is he snapped one of their down hill frames years ago, then when spoken to, said he knew what had caused it and could snap them on demand if they wanted him to, this led to the association and the development of improved frames….or something like that.
Frames shouldn’t break, cars shouldn’t crash and houses shouldn’t catch fire but they do.
Really glad BK is ok.
Pivot shares, sales, confidence in products!!!
Bottom line is this particular bikes are getting pushed to a point/level that 99.9% will never get near! So confidence shouldnt be lost in Pivot frames
www.youtube.com/watch?v=XJ-G0p9bnFY&t=182s
I don't care how much surface treatment you give the aluminum lugs, corrosion will happen.
But, BK and the rest of the pro's are "test pilots" of sorts, they are paid to push these materials to the brink. Looks like BK found this prototypes limit!
Why does the manufacturer even take this risk for potential negative impact on their reputation? I guess nobody did a 'what if...'
Maybe like that?
www.instagram.com/p/B4NdjEnohIi/?igsh=MXQ2bGQ0M2p5azB5dA=
It’s a prototype
And Has there ever been a production bike failure from all brands before?
GAL
"Bernard likes the pies".
This is the real reason why his frame snapped but keep it between ourselves.
- different materials with their different thermal expansion coefficients creating relative movements and resulting in forces in all directions
- the different stiffness of the lug and the tube in the overlapping area due the material properties and the form factor (design) to prevent stress raisers in all areas and to minimise relative movment during deflection to reduce the shear inside of the adhesive film and in the connection faces between the adhesive/tube and adhesive/lug
- The manufacturability of the lug/ the tube. Not all shapes can be produced with the chosen process like the cnc milling of the lug and the carbon fiber tube lamination.
- The assembly order of the tubes, especially, when there is a radial overlapping between lug and tube. The last lug must be able to be assembled in one direction. Like already stated by WRCDH the tube directions at the Head Tube Lug should have different directions to allow for a form-fitting. Apparently the Pivot engineers have decided to use the head tube lug as the last bonded lug. This assembly order however prevents the desired form fitting in one of the most stressed areas of the frame. During a touch down after a jump, there are huge forces, that tend to pull the lug from the tubes - just have a look at the „hug to flat“ videos and the deflection of the fork stanchions in the driving direction! And obviously, this is the direction where the lug is completley reliing only on the adhesive connection working properly, there is no redundancy here, when the tubes and their assembly direction of the overlapping area of the tube and the lug is arranged in parallel and in the force direction like on the Pivot. The Pivot Enginners have tried to minimise the risk by incresing the size and the length to diameter ratio of this connection, which is obviously not enough.
As you can see, all of these design parameters are often contradictory and in fact even more often contradictory to the intended behaiviour of the whole system like the frame or the whole bicycle in the field (overall frame stiffness vs. connection stiffness of the lug, etc.).
This is quite challenging if its a design like a prototype dh bicycle, where the amount of financial investment, that you are able to put in is really limited and at the same time you have a vast amount of different and confusing influences from manufacturing (bonding surfaces preparation, curing temperature, correct adheasive bead thickness, al these parameters must be obtained by a huge number of repeatable experiments, inputs from the adhesive supplier, etc.), human behaviour and imperfections during the assembly - we are not robots allowing for 99 % of repeatability, a huge type of hardly predictable external impacts like different strains (e.g. multidirectional loads, temperature changes, etc.) - this data must in general be collected by data aquisition in relevant operating states and environments: e.g. on different tracks (min 3) with min. 3 different riders in advance of the design process.
During this process you need to execute comprehensive lab testing with your saples, taking into account the relevant load cases, This testing must show a sufficient life span, prior a testrider enters the bike.
Most important: at the end there must be a replicable „closed loop“ between all these Influences and Variables and your design of the product and your manufacturing and assembly process and the user instruction.
You need to be clear about the fact, that, as an engineer, you are responsible by law to control all these influences and even for a prototype (!) using all currently known and available methods of the so called state of the art, to prevent such a katastrophic failure like happened now.
Just the fact, that a lot of bicycle prototypes regulary fail, does not liberate you from these obligations. And if you are not able to prevent a failure by the design (over the entire lifespan and a reasonable amount of abuse by the user), the process control and all its parameters, the product design must be executed in such a way, that the failure is „anouncing“ itself for reasonable amount of time e.g. and amount by visual deformation, unusual and massive sounds etc. - and not happening instantly.
All these steps, measures and countless (design) desisions needs to be documented thorougly in these days. Even for a prototype! There is a chance that you will be forced to show and explain this documentation in detail in front of a judge, if something goes horrible wrong. You need to always keep this in mind as a responsible Engineer/ Engineering Manager or Managing director. But on the other hand, it should not prevent you from beeing innovative and find new solutions. Very often this is a hard fight in engineering departments, or between departments of product centered companies in general. This is not an easy Job.
What really astounds me is, how fast pivot has drawn its conclusion - thats seems to be to fast for my understanding.
From my point of view all these kind of frames should be immidiately taken out of operation, the analysis should be executed thorougly in house and in parallel by an accepted institution like an university or an independent testing laboratory. Based on these results - and only then, and after a thorougly process and design considaration and a comprehensive risk analysis, they should take and announce countermeasures. But maybe I am wrong and they have done all that…
my guess is that all the boutique brands that wanna go to the big dances are paying attention to this incident
Titanium: By looking at the standard electrochemical potential of titanium, it seems that this metal is an active metal. However because of the formation of a dense stable and protective oxide layer, titanium is placed among the noble materials and just below graphite or carbon in the galvanic series table. (For a primer, see the article An Introduction to the Galvanic Series: Galvanic Compatibility and Corrosion). Therefore, there is no significant gap between titanium and carbon-fiber-reinforced composite to create galvanic corrosion. This means that commercially pure titanium and its alloys are completely resistant to galvanic corrosion when they are coupled with carbon composites.
www.instagram.com/reel/C4r8uvvvGbi/?igsh=N3didzAzbDd5N3cx
www.youtube.com/watch?v=aCMU9xebt7Y
What's odd is that it takes a bunch of hits, then right at the end as it's coming to a stop it separates into 2.
Glad this thing didnt get pulled apart at hardline then
On a serious note, cable routing may be an issue then…
True about the cable routing! Surely there’s a way of keeping the down tube in compression? @rich-2000:
So maybe 60-100 lbs of suction force, vs easily 20,000+ lbs of bond strength.
Wasn't one enough?
Prototype 1: Include head tube pivot
Prototype 2: Don’t include head tube pivot
scroll down, no conspiracy here
"...more questions than answers. Why did the tubes slip out?"
The epoxy did not bond to the aluminum. Did you read the explanation?
I work and have worked with epoxy a lot in construction (concrete foundation hold-downs, slab countertops, etc.) and with golf clubs. Epoxy is extremely strong and aherent.To me, there is a lot more questions.
Why do you care what I think? Have a better comeback than "username checks out." Lmao!!
You keep posting comments to support my argument. Lmao!!
…..my mind
The first flaw is the straight-insertion of the top tube AND the down tube into the lug, at the same draw angle. As such, if the bond isn’t perfect and the bondline fractures, the tubes are not mechanically constrained by different draw angles...both tubes pull out of the lug(s) in the exact same draw direction — that is a serious design / engineering flaw in my professional (composites engineering) opinion. If designed differently, with different draw angles for each tube / lug, both tubes likely would not experience the same stress vectors at the bondine (due to their varying orientation) and thus both bondlines likely wouldn’t fail at the same time. And both tubes would not (and could not) pull out of the lug(s) at the same time, in the same direction. With different draw angles / bondline orientation / stress vectors, one bondline would likely fail before the other, allowing detection of the bond failure. Likewise, the different draw angles (if the tube are stiff & strong enough) wouldn’t allow both tubes to pull out at the same time, in the same direction. Atherton got that right. Pivot got that wrong. It was a major oversight and a fundamental engineering flaw, in my opinion. That Pivot frame even has a separate individual top tube, and a separate individual down tube — so they didn’t need to utilize the same draw angle for both the top tube & down tube. (But the same draw angle would be required if the rest of front triangle was a single-piece composite molding and the top and down tubes weren’t separate pieces).
Also, if the bondline wasn’t perfect, bonding failures can occur from bonding adhesive microcracking, adhesive macro-cracking, and incomplete adhesive coverage. Galvanic corrosion, and other corrosion, can be an issue with aluminum lugs and carbon fiber (unless the carbon fiber is insulated with fiberglass or similar non-conductive material). Aerospace companies often use fiberglass layers to prevent galvanic corrosion, making sure to not sand them off, nor sand too deeply into the fiberglass, where it bonds to aluminum. Likewise, at a minimum, they’ll typically hard anodize the aluminum, or use another bonding-optimized mil-spec surface preparation or treatment and/or anodizing process. Often, aerospace companies apply a military-spec type of cured epoxy primer onto the aluminum, to which the subsequent prepreg composite, or bonding adhesive, can tenaciously bond (and bond without corrosion due to the protective anti-corrosion epoxy coating). It appears Pivot excluded some or all of these critical (and fairly simple) steps. I also don’t see any fiberglass on the cured tubes, in the bonding area, in the prototype manufacturing pictures that Pivot released for publication on Pinkbike — it appears to be solely a woven carbon fiber surface layer, with no galvanic insulation layer, in the area bonded to the lug.
Additionally, it appears the bonding process was flawed — I didn’t seem to see much adhesive on the composite nor aluminum portions of the failed bond (in Kerr’s broken frame pictures). There’s a chance they didn’t properly remove mold release on the cured composite tubes. Or that they didn’t properly prepare the composite & aluminum surfaces (via abrasion and solvent cleaning), or that insufficient bonding adhesive was used, or that the adhesive scraped out of place during the assembly process...thus resulting in insufficient adhesive in the area to be bonded. The bonding process also appears suspect, from what I’ve seen in various pictures of the frame & manufacturing process. Trek (OCLV) and others co-mold several 0.002” tall “pips” around the circumference and length of their “male” bonded tubes...those pips center the tube into the female lug (or tube) and ensure an optimal 0.002” thick bondline, and it ensures that the adhesive doesn’t get scraped out of place as the tube is inserted. I doubt Pivot did anything like those 2-thou pips, based on several of my observations. It’s also usually a good idea to “inject” a bit of structural epoxy bonding adhesive into a port at the bonded joint, after initial assembly with adhesive, to ensure the entire bondline is fully covered, saturated, and void-free. I don’t see that sort of adhesive-injection port feature on their frame (from my cursory review), but I think Specialized might do that on their prototype DH frames. I think Atherton does the same. It’s common in aerospace, F1, and even in the production of certain composite sports cars. I actually import special 3M structural adhesive from the UK (it’s what I used on McLaren’s carbon fiber “tubs” and Bentley Motors’ prototype carbon fiber doors for one of their Super Sports technology demonstrator cars) which is better than what’s available in the US (likely because it’s more toxic =) — it’s strong, and has superior mechanical & dynamic performance numbers, and is highly toughened (as well as being highly crosslinked for extra strength, if you post-cure it a bit around 250F to 350F after achieving initial bond strength at room temp, overnight. Engineers, prototype manufacturing technicians, and production manufacturing technicians, cannot mess around with bond failures — it has to be done right, and the materials and manufacturing processes have to enable bonding perfection. Every time.
Also, Pivot appears to use cast-silicone custom bladders for their prototypes — and silicone / silicone oil can readily leach into the prepreg (unless significant protective countermeasures are taken). The silicone contamination can greatly adversely affecting bonding / adhesion. Aerospace companies ban everything silicone in their composites manufacturing facilities, for this reason. Even those yellow silicone “Livestrong” Lance Armstrong bracelets are strictly banned. Pivot’s custom composite tubes could have been contaminated with silicone that could affect their adhesive-bond performance — and their manufacturing pictures show the silicone bladder directly in contact with the uncured & cured carbon fiber prepreg, directly adjacent to the end of the tube (where the composite tube is subsequently bonded). Polyurethane, latex, molded nylon, or nylon film over polystyrene foam rigidized-bladder mandrels, are good alternatives for prototypes (and sometimes production).
Lastly, the kink / bend in the carbon fiber downtube, just behind the lug, will put both sides of the downtube in compression right at the kink / downtube bend...and carbon fiber fails in compression. Although Kerr’s bike didn’t fail there, it’s a suboptimal compression-loading scenario which will result in pronounced compressive forces in the laminate on both the top and bottom sides of the downtube at the bend (in a jump-case loading scenario like Kerr experienced). Frames have to be engineered, in every way, for those worst-case cases...and engineering a frame with a tube geometrical configuration that results in substantial compressive forces on both sides of a composite tube is just questionable engineering, in my opinion. The goal should be to limit inducing those sort of compressive loads and mitigate the risk of composite compressive failure.
-WRCDH (guy above)