MTB Compliance Analysis: Danny Hart's Modified GT Fury Frame

Danny Hart's GT Fury caught a lot of attention this past World Cup season as it had some notable structural frame modifications to increase compliance. Dale, Danny's mechanic, is a veteran frame engineer and had some great ideas on how to fine tune Danny's bike to work best for his style and to eke out the literal hundreths of a second needed to win races. If you haven't seen these mods, check out this quick section of Vital's PIT BITS video here. 

Let's dig into each modification. This will be extra fun because I was fortunate enough to be the lead engineer on the GT Fury frame, so I'm pretty intimate with how and why it works. My first notes are that Danny isn't a huge guy. He's 5 ft 10 in (177.8 cm) and weighs 154 lbs (11 Stone). The Fury was designed with compliance in mind, and learnings from previous generations of frame designs before it led to where we wanted the frame to be stiff and where we wanted our compliance to come from.

Our approach with the Fury mostly aligns with what was discussed in my last compliance article. TL;DR - A bike seems to handle more consistently when you introduce compliance through the seat stays, allowing for more roll while keeping stiffness in the chainstays and between the connection of the chainstay and the front triangle, reducing wag. This roll-biased agility allows for frame compliance with less compromise in handling. It ensures the bike maintains good pedaling stiffness so that it doesn't flex under pedaling, reducing the perceived pedal efficiency.

Combining this roll-biased compliance with a reasonable amount of anti-squat makes for a race bike when pedaling out of the gate and out of corners. You can have high anti-squat, but if your bike is a noodle, it will not feel as peppy as a bike with high anti-squat and stout pedaling stiffness, where the bottom bracket doesn't wag side to side under hard pedaling. It may not actually be faster, but it feels faster. Feeling fast is great for your head, and your head can be why you win or lose races.

It makes sense that Danny would want more compliance if he's not a big guy. It also makes sense to design the stiffness of a bike to accommodate the majority of riders and compromise for as few as possible. If you design the stiffness for the average rider, the bike may not be stiff enough for all of the riders heavier than that average rider. If you design the stiffness for 80% of the riders, it's a good balance where most riders will get enough stiffness, but you still have some flex so that all riders aren't miserable riding the bike. 

Slight Detour to Talk About Frame Manufacturing

You may ask, why don't you make a bike perfectly stiff for every rider, right? That's the ultimate goal, but it's never that simple. You often have to use the same frame parts between all frame sizes because if you didn't, the cost would be so high that people wouldn't be able to afford the bike, and then you wouldn't have a business. If you don't have a business, you don't get to make bikes… and that's no fun.

I know a lot of people get this, but it's worth mentioning as a reminder. The engineering challenge is often: how do you make a bike that makes the most people happy while also sustaining a business? It's not as fun as making the world's fastest bike at all costs, but this isn't F1, where the racing IS the business. Most bikes need to win on Sunday and sell on Monday, and that's okay. 

I would argue the stiffness of the Fury is a great balance of enough stiffness for most people who ride it and still provides enough compliance to keep people comfortable but ensures there isn't too much compliance for the larger riders, which would hurt their handling. It is biased toward being a little more stiff than a little less stiff for this exact reason, and I believe that if given only one chassis to work with between all frame sizes, this is the best balance.

The regions highlighted in white show where material was removed. 

We see the modified parts: the link, the seatstay bridge, and the chainstay forgings. So, why modify these parts and not others? Practically speaking, these are the only components you can change without creating entirely new parts, which often take too long to produce and cost a lot of money.

For example, if you were to design a custom-machined seatstay bridge, you'd need to design it, send it out for machining (where it joins a queue with other projects), and then ship it back to weld it to the seatstays. Coordinating the seatstays requires working with your busy manufacturer, welding, and then heat-treating the part. Where's the closest heat treatment facility? Can you join their next batch? Heat treating just one part is too expensive, and most bike companies don't have their own heat treatment oven. This one part could end up costing many thousands of dollars but more expensive is the opportunity cost of not working on production projects that provide revenue for the business. Every step is surprisingly expensive. 

You see where I'm going with this—it's a lot of time and effort. And often, you simply don't have that time because the first race is next month, and you still need to test ride the bike before then. The solution? Work with what you've got! 

How To Find the Right Amount of Compliance

Step 1 is to add as much compliance as possible to see if the bike can meet the rider's preferences or provide the feel they're looking for. It's essentially a form of bracket testing. Let's dive into each of Danny and Dale's modifications, discuss its purpose, and review a simple Finite Element Analysis (FEA) model to measure how much these changes affect stiffness.

Red shows where the most movement is happening when this FEA model is loaded.

This FEA model I put together is pretty simple and just focuses on rear triangle flex. Yes, it should be at sag… actually, past sag for high G compressions, but this is more of a quick taste of what the mods do to the frame, so let's not worry about it 😊. A lot of people have never gotten to see a simulation like this, so it's fun learning for all. The model has a fake rear wheel so we can visualize any wag and roll. The red coloring shows where the greatest movement is happening relative to the center plane of the bike. (If anyone wants to see a specific mod and its effect on this frame mock-up, let me know, and I will try to run it.)

I used 1000N applied to the contact patch of a 27.5-inch tire (defined as rigid in the model), pushing to the rider's left like you were in a left-hand tight berm. The load and these displacement numbers are somewhat arbitrary and are more for A-B comparison between the mods. I chose this load because I needed a load that would flex the frame enough to allow us to easily see a difference between each mod. Below, we can see the unmodified Fury frame being flexed in the FEA.

The stock, unmodified frame has the rear wheel contact patch move 11.5mm relative to the bottom bracket. We can see the rear wheel tilted to the side in the rear view, indicating roll. The top view shows the hub has shifted over to the side and is now at an angle, indicating some wag. Note these displacement images are exaggerated by a scale of 4X to allow you to see the movement more easily so take them with a grain of salt.

Link Modification Analysis

Let's start with the modified link. The stock is wild in itself. It was made to look cool and compliment the Fury's unique frame layout. It's not structurally efficient, but that's okay. It ended up beefy, but it's in the center of the bike, so you don't notice it much. We see plenty of World Cup teams adding weight to their light DH frames, so let's just say this bike has that centralized ballast built into the frame, from the factory 😉 Regardless, you can cut some material out of the linkage, and it could still be safe to ride depending on how you do it. However, mods like this can be strong enough for now, but the fatigue life of the part can be dramatically reduced, and in a year or two, you may see a part failure. So please don't do this to your frame. Dale did it safely, using his experience in bike linkage design and with this being a race bike, it's only going to be ridden for a season, max.

Now, the "web" between the two sides of the linkage is removed, and the seatstays can flex up and down relative to each other, allowing for some roll out of the rear triangle. 

Seatstay Bridge Modification Analysis

This brings us to the seatstay bridge modification. Let's say you didn't modify the link and only reduced the size of the seatstay bridge. The link is stiff enough that not much compliance would be induced, even with the smaller seatstay bridge. The same applies if you modify the link but leave the seatstay bridge untouched, though to a lesser extent. I could show you the comparison in FEA, but these models take time to run and troubleshoot—so just trust me on this one; things change, but not much.

Modifying only one part will add some compliance, but "some" isn't that useful for learning. More than "some" is. By modifying both parts, you can achieve greater flex and fully experience the pros and cons, knowing you can always revert back to an unmodified part later if it becomes too compliant. 

You could remove the entire seatstay bridge, but this could cause problems over time. Without the bridge, the bearings pressed into the link might start to walk out of the link, as the original design relies on the seatstay bridge to keep the link-to-seatstay assembly aligned and the bearings securely in place. Fully removing the bridge could also cause the seatstays to contact and rub against the link. Dale wisely retained a small portion of the seatstay bridge to avoid this issue—good call. 

Link + Seatstay Bridge Mod Analysis

Modifying both the link and the seatstay bridge introduces additional roll to the frame compared to the unmodified frame. As shown below, removing the seatstay bridge and modifying the link allowed for increased compliance.

The modified seatstay bridge and linkage cause the rear wheel contact patch to move 14mm relative to the bottom bracket. We can see there is more roll compared to the unmodified frame. The amount the hub moves to the side looks fairly similar to the unmodified frame, indicating we see more roll introduced instead of a wag.

Chainstay Modification Analysis

Like the link, the chainstay design is unique due to the aesthetic of the Fury. The goal was to keep this part stout to maintain high perceived pedal efficiency and ensure the bike doesn't wag in corners. However, if your bike is already stiffer than you want, you weren't going to experience wag in corners anyway. Adding compliance that gets you closer to introducing wag isn't a big deal, as the tracking and comfort of the frame can often be more important.

The chainstay tubes themselves are fixed—they would be very difficult to modify—so once again, you have to work with what you've got. Removing material from the chainstay forgings makes sense. Why there? Dale explains this in a few videos, and his reasoning is spot on. Removing material from the bottom of the chainstay forging is the safest way to add compliance without increasing stress in the areas that see the most load. Doing this modification alone, without other changes, would introduce more wag than the stock Fury. However, combining this with modifications to the link and seatstay bridge creates a frame with a similar roll and wag to the original design but with a notable increase in compliance.

With all mods applied (seatstay bridge, linkage, chainstay forgings), the rear wheel contact patch moves 22.1mm relative to the bottom bracket. We can see from the top view that the rear hub has shifted quite a bit over to the side and is angled, indicating wag. This also means the bottom bracket can more easily move side to side when pedaling.

So what did we see in terms of differences in compliance between the modified and unmodified rear triangles? We can go into percentages and numbers, but since this isn't the real model of the rear triangle and just a mock-up, they would only be so useful… but we'll do it anyways: 

It's more compliant now. I'm not sure where "too compliant" is for Danny's speed and weight. Maybe he will let us know? (Danny, if you're out there, let us know🤞). This data is only useful with rider feedback but it's fun for this article. 

One last mod to talk about—and this one is neat: the endcap for the hub, designed to keep your hub from moving much relative to your brake mount. This often doesn't get much attention. If the brake mount area of your frame is super flexy, your rotor will push into your caliper, causing it to drag on the pads. In extreme cases, it can even reset your brake positions during hard cornering—especially if you're dropping into a corner so hard it feels like you're landing sideways… or if you're actually landing sideways. This is known as brake pad knockback.

I personally suspect this is one of the causes of Shimano's wandering bite point: a stiff rotor combined with mineral oil-lubricated pistons (slippery), plus a compliant frame loaded in a corner, can slightly reset one side of your brake caliper. This requires you to pump the brake to get it back to normal. That said, this doesn't happen with all brake setups, but it's worth considering in frame design. Some brakes might get a bad rap because of a flaw in the frame. Like how shocks sometimes blow up, and people blame the shock company, when in reality, the frame wasn't even close to aligned throughout the suspension travel.

Back to Danny's bike. I've seen bikes where the dropout is so flimsy you can grab the rear wheel and cause pad knockback with an uncomfortable amount of ease. The Fury doesn't have this problem, but at the high speeds and loads of a World Cup rider or track—combined with large, stiff 200mm+ rotors—it might happen occasionally. Once you're aware of this, you might as well do everything you can to reduce or avoid it. That's what Dale has done with this dropout endcap. Most riders will never experience this issue, so no, you don't need to rush out and have a custom endcap made. And since he's adding more compliance to the frame, pad knockback becomes more likely, so it's a smart move. It's like the Torque Caps on RockShox forks, but for the rear. It's cool, and it's a great idea. More brake companies may want to set a minimum dropout-to-brake-mount stiffness requirement for frame companies so their brakes don't get blamed for issues caused by frames. But that's up to them. It also might not even be a real problem and I could be completely wrong.

Anyhow, hope you liked this article. Let me know what questions you have and fun things you want to talk about that relate to this. We can dig into this simple FEA model and try some things on it, happy to nerd out. If anyone is making their own bike frame and has it modeled in SolidWorks and wants to look at it on FEA, hit me up via DM. It could be fun to run an FEA on it and make recommendations on how to get more out of your frame design, and we can all learn and get better from it. Shred on and Happy New Year!