Epic REACT Mods and Carbon Fiber experiments

[Four by Six]

I’d like to detail some experimental modifications I performed on my Epic REACT shoes. I wanted to work from smallest to biggest changes, in case I messed up and ruined the shoes.

1. First, most obvious mod, taking off the shoe laces. With the full flyknit upper, they don’t do much anyway. I found they do not help foot lockdown much, and if you do need the laces for lockdown, either you bought your shoes too big or have running form issues.

2. Remove shoe liners. This greatly enhanced the comfort and made the slip on like, well, slippers. The foot did become more loose, not so much I could not run in them, but it might cause some people blisters. After a few days of wearing them, my heel started to hurt somewhat.

3. Removing the outer lace hole reinforcement. The shoe looks a bit goofy with no laces but lace reinforcement panels. There is no issue with loss of reinforcement, as the main reinforcement is a full welded panel in the inside that includes the swoosh. The outside is just glued on cloth. On one shoe I used solvent, which worked well. On the other shoe I used brute force pulling, and it left a layer of glue behind that needed to be dyed black.

4. Cutting off midsole flare. I noticed that the bottom of the sole flairs much wider than the shoe. I assumed this was for stability and not just for show. It bothered me a bit, and sometime felt like I was wearing clown shoes. I took a sharp knife and just cut away the flair, making the sides square with the shoe. To my surprise, after several days of wearing, including running, I came to the conclusion the mod did not impact stability at all. My feet are just as stable with front, mid, or heel strike landings.

5. Creating artificial heel. I noticed that the shoe feels banana shaped on the bottom, which helps with heel to toe transition, but also bothered my arch during casual wear. I took a sharp knife and cut away some of the middle of the foam outsole, thus creating a small heel where the plastic heel counter is located. It seemed to relieve my arch pressure, but put additional pressure on my heel.

6. And finally, the craziest modification of all. I noticed that the bottom of my foot got really hot, especially when running longer distances. This surprised me, as the shoe is very breathable. So, I drilled 1/4" holes in the sole, from front to back, careful to keep close enough to the middle to not hit any of the stitching that holds the upper to the midsole. As the holes tended to self-close since the drill does not work with foam very well, I used a hot soldering iron to melt the jagged holes into more open rounded holes. This solved the heat retention problem, and it really feels strange having cool air hit the bottom of your feet when running. Much to my surprise, the REACT foam has much more bounce now, as there is more weight per square area of foam, as I have cut and drilled much away.

Anyway, crazy to do all this to a new pair of expensive shoes, but they are still not only usable, but runnable. The only mod that impacted running ability was creating the low heel, but even that is bearable. My shoes are now much more comfortable, cooler, and lighter. (I weighed the size 10 pair as 400g even.) As I use these for triathlons, they pull on very fast in T2. And in the summer heat, on sun soaked roads, my feet stay cooler.

 

[Four by Six]

Here is a pic of the holes in the soles. I realize I went overboard, but I did it in stages. In the end, the ventilation and water evacuation is fantastic, but I removed too much REACT foam, and the midsole is now like jelly, squishing and jiggling everywhere as I walk/run. Very comfortable as a slipper, but not stable enough for a running shoe.

I did try a few more mods, before calling it quits.

7. The shoe bed is not a single layer of cloth glued to the REACT foam midsole. It is a sandwich of white felt and a fiber reinforced tape. The felt slips on this tape, causing hot spots, as noted by many reviewers. The felt also holds water. So, I cut the felt off on one of the shoes. In the picture below you can see the fiber reinforced tape as it has a square pattern printed on it.

8. My heel pain persisted, so I reinstalled just the heel portion of the insole. I cut the insole to line up with the outer heel counter. This greatly improved heel cushion, as well as returning the offset, as the shoe almost feels like a racing flat without the insole.

9. Finally, my last mod, I removed the inside lace/swoosh reinforcement panel, which I at first thought was not possible. This made a huge difference in the comfort and flexibility of the flyknit upper. After removing the panel, I stretched the knit upper various ways to loosen it up. With all the panels all over the flyknit, it wasn't really able to stretch much in the mid foot. Now it feels amazing, so much so, I think you could leave the insole in and have enough stretch to not have to remove it or cut it down.

So, after all these mods, and wearing and running in the shoes, I think the only real useful mods worth repeating is the uncaging: removing the laces and lace reinforcements inside and out. It really takes the shoes to the next level. Cutting the flair off the midsole doesn't hurt, but doesn't help either. Forming a heel put too much pressure on my heel. Drilling holes in the sole made the midsole too unstable, though the ventilation was nice. Removing the insole isn't necessary if the flyknit is allowed to stretch more without the laces. As for the white felt slipping, supposedly Nike is glueing these better now, but if not I will make a slit down the middle and glue it myself.

I have another pair being delivered tomorrow and I'm going to try and see how cleanly I can uncage them. Perhaps since Adidas eventually offered both uncaged and laceless versions of the Ultraboost, perhaps Nike will eventually follow suit and offer a laceless version of the Epic REACT.

 

[Four by Six]

OK... Couldn't resist one more mod...

10. I removed the outer heel cup. There are both inner and outer heel cups. The outer is plastic and the inner is like a synthetic suede. There is more flyknit under the outer heel cup so I was curious if the shoe would be a bit more comfortable without the outer heel cup and whether or not it is even needed. After removing the outer heel cup, the shoe became more comfortable and I like the look, but the heel is really unstable now. I think the heel is stable enough for use as a slip on leisure shoe, but not running. Evidently the inner heel counter is not enough to hold the heel in place.

 

[Four by Six]

My new black/black Epics arrived, and I really thought long and hard about what mods to do and not do, as I didn't want to ruin these shoes. While I originally planned on making them laceless, I decided to keep the laces and lace reinforcements and instead remove the swoosh and swoosh backing. I didn't remove the swoosh for looks, and you can still tell there is a Nike swoosh there. Rather, the swoosh and the reinforcement panel behind the swoosh impede the flexibility of the flyknit. Also, the swoosh backing material is used to connect the lace area to the heel counter. Removing the inner swoosh backing frees up the flyknit to do its job better.

This mod makes a HUGE difference. The Epic is typically very snug in the top forefoot area. Some owners are even removing the insole to give their foot more room, or going up a half size. I was stunned at the difference in just removing the swoosh backing material. Evidently it is this material that locks the forefoot down to the shoe and connects the lace area to the heel counter. With only this single modification, the shoe went from too snug in the middle to extremely comfortable. No need to remove the laces or the lace hole reinforcement.

I'm probably going to paint the blue heel counter black, to have a true triple black, but that is all I'm going to do to this pair.

 

[Four by Six]

Per request, I am uploading a pic of where I made the cut to the rubber strap that forms the inside backing of the swoosh. Sorry, I don't have a before pic, and I already threw away the part I removed, so no pics of those. It is pretty straightforward, though. You make the rear cut along the heel cup, and then the front cut along the lace reinforcement, and then tug with a pair of pliers to rip it off. I then stretched the flyknit with my fingers to get it stretching in all directions.

 

[Four by Six]

My Angelus paint arrived, so I painted my black Epics to make them true triple black. I used flat black paint on the blue heel counter and pull tab text. Here are two views of the after results...

 

[Four by Six]

Since I essentially ruined my OG pair of Epics, I ordered a replacement pair to make my everyday casual shoe. And since it will be an EDC shoe, I decided to go for broke and fully uncage the pair.

I took a pic after removing everything but the swoosh, so you can see what it looks like with the swoosh. I like the looks better with it, but the flyknit is a tad more flexible without it, so I went ahead and removed it. The heel cup says Nike anyway.

Here is a pic of everything I removed, per earlier in the thread: Laces, outside lace reinforcement, inside lace reinforcement, outside swoosh, and inside swoosh reinforcement.

Finally, here is a pic of the end result.

I could still remove the outer heel cup, as there is black flyknit under it. But that made the heel area a bit too unstable. Or, I could paint both the heel cup and clip black, to make the shoe more Oreo. However, I kinda like the added splash of color (for now).

My perfect pair would match this flyknit upper with a black insole and black heel cup/clip, but there is currently no way of creating such a combination.

 

[Four by Six]

I thought it would be crazy to run in this Moc version of the Epic, but I've done just that the last few days out of curiosity and it worked great, much better than I expected. This modified version still fits tight enough to stay on the foot in a fast run, yet it was flexible enough to let your forefoot flex and toes spread on landings. While I would still prefer my laced black version for races, the Moc fit snug enough that I don't think I have to worry about twisted ankles or shoes flinging off.

Some other Epic experiments I did to my OG pair before I removed the upper completely:

I gradually sliced thickness off the mid/outsole making less heel drop by shaving a few mm off heel taper. This felt amazing and forced a bit more forefoot strike, and my heel did not drag late in the run when my legs got tired. Next I kept the lower offset and just removed more REACT foam front to back. Again, this felt amazing. There was less vertical displacement cushion in the midsole, you could feel less "give", but there was still just as much rebound and bounce. Felt GREAT. Finally, I tried making a racing flat by removing all but just 1cm of REACT foam from front to back. This felt OK, but with no plate, there was not enough structure to the bottom of the shoe, and it felt like running in Vibram 5 fingers - you felt everything under the foot. I bet with a carbon fiber insole, it would have felt great, but I did not have one to try. Finally, I just removed the upper to try on a different midsole in the future.

 

[Four by Six]

I'm still trying to perfect the "runnability" of this shoe - it has so much potential. While the REACT foam has some structure to it, primarily due to its thickness, I thought the shoe needed a little bit more. The strobel board is really just fiber reinforced tape - what the fibers are I have no idea, but it is not very stiff. I wanted to try a carbon fiber plate in these Epics, but I haven't been able to find what i'm looking for. Everything out there is a flat plate for foot issues. And I don't want the ice cream scoop shaped plate that throws you forward in the 4% or ZF. I just want something that sandwiches between the midsole REACT and insole that is shaped with the same curves as both.

Since I could not find what I wanted, I attempted to make one. To get the right shape and not mess up the shoe's dynamics, I masked off bottom of the insole and top of the strobel board. I then took casting tape and put a layer between them, put on the shoes, and proceeded to walk around until the casting tape began to set. I then removed my foot and let it set up hard. Once it was set firm, I removed the casting and applied layers of weave I had laying around, one with fiberglass and another with carbon, and let the resin cure.

Results? They both feel good. The fiberglass is too thick and heavy. The carbon felt great, but it is a cross weave pattern that does not provide enough front to back stiffness and springiness. I'll have to order some carbon tape with linear fibers and run it front to back. However, for now, the shoe feels much better with the fiber plate between the insole and midsole. Just enough additional structure to let the REACT do its job, without too much stiffness either.

 

[Four by Six]

I'm done making the new set of carbon fiber plates that will sandwich between the insole and REACT midsole. The casting turned out great with only a couple small wrinkles in the left heel section. These are much stiffer, but I hope also flexible as well. I don't have a vacuum chamber or bagger, so the resin is just wiped on with a brush, thus the quality is not on par with factory made CF.

They feel great in the Epics. Once I have a few runs in them, I'll comment further on performance.

 

[Four by Six]

Once I have a few runs in them, I'll comment further on performance.

After a 10K run last night and my morning 5K, I have feedback on the plate's performance.

First the good news... Performance was off the charts! Before, the REACT foam had plenty of cushion, but I never really felt any "energy return" like I do with TPU and PEBA foams. With the plate, I could feel the REACT foam springing me forward. It was not the spring of the plate flexing. It was having something firm to press in the foam and release. A foot by itself is just too soft to compress the foam correctly for energy return. That must be why Nike puts a plate in the ZF and 4%. Seriously, the feeling was amazing.

Unfortunately, the feeling did not last. By the end of my run last night the plate started giving in and lost its rigidity. While carbon fiber does not fatigue, when used with too much resin and air bubbles, the composite will fatigue. Also, my stack of casting tape and carbon fiber was too thick and made the upper too tight, cutting off some blood flow and becoming uncomfortable by the end of the run.

So, I went from elation to disappointment. I still think a carbon fiber plate molded to your foot is the ideal solution for the Epic. I will have to find someone with a vacuum chamber and/or a vacuum bagger for CF lamination. I also might try CF prepreg, but that requires 350F to cure and I think the Epic uses heat activated glue to hold everything together, so the shoe might fall apart after heating. Doh!

 

[thebtcbtq]

I am interested in carbon fiber insoles. Looking forward to your next post!

 

[Four by Six]

While I wait for my CF supplies to arrive in the mail, I have been practicing my resin technique with fiberglass cloth as it is very inexpensive. So far all my attempts have been a disaster. Thick, uneven resin with voids and air bubbles everywhere. I've tried several techniques and types of resin so far. The only way of getting a good resin job is to use a rigid mold, perforated release film, and vacuum to draw the resin across the fibers evenly and suck out air bubbles. That is more work and money than I want to expend. I'm hoping to find a cheaper/easier way.



That's the bad news... The good news is one pair came out somewhat useable (though a little thick and heavy) and I set a new PR for my 5K neighborhood loop using them by a surprising margin. And afterward my body did not feel beat up at all, though I definitely felt the additional weight of the shoes. Unfortunately, the run destroyed the plates as I found cracks all over them afterward. If I ever manage to make a decent CF pair (thin, light, springy), I'll alternate with my control shoes (Pegasus 34) and compare the average times after a bunch of runs. Or I guess I could alternate with and without the plate in the Epics and see if the plate really makes a difference or if it is just a case of confirmation bias.

 

[Four by Six]

Running the neighborhood 5K loop this morning without the now cracked plates, I felt fast and thought I would match or even beat my PR set yesterday with the plates. I was wrong. My time matched my PR set before without the plates. The plates were over a minute faster, though admittedly this is only one data point. I'll have to take more data in the future when I have a new set of working plates made.

 

[thebtcbtq]

Running the neighborhood 5K loop this morning without the now cracked plates, I felt fast and thought I would match or even beat my PR set yesterday with the plates. I was wrong. My time matched my PR set before without the plates. The plates were over a minute faster, though admittedly this is only one data point. I'll have to take more data in the future when I have a new set of working plates made.

It sounds like the CF plates absorb more down force energy. That is good to hear. Hopefully you can make some progress on the r&d issues, I know that working with resins and composite materials is not easy.

 

[Four by Six]

It sounds like the CF plates absorb more down force energy.

Not sure the reason. The nearest I can tell, for the same level of effort and exertion it appears each stride is approximately 1-3" longer. I run on sidewalks with evenly placed lines, and it appears I am covering a bit more ground per step. Of course, with the disclaimer I only have a few data points thus far, and I'm certainly not running in controlled laboratory conditions. On the other hand, I keep getting 1 minute saved on a 25 minute run with the plates versus without the plates, and that works out to... 4%. Where have I heard that number before?

 

[Four by Six]

I know that working with resins and composite materials is not easy.

It definitely is frustrating. Just getting the carbon fiber takes forever. I guess I'm spoiled with Amazon Prime. You can only get quality carbon fiber materials from industrial suppliers, and they certainly aren't in a hurry to fill the tiny orders I'm placing. Carbon fiber is expensive, and most of the good stuff has large minimum buy orders. I ended up placing small orders from several online suppliers and I'm still waiting for most of it to arrive.

In the meantime, I have been playing with carbon fiber stackups with what has arrived already. To save costs I have been experimenting with half plates, which are a close approximation to the full plate. Once I figure out a preferred stackup, I'll make a set of full plates.

Below is a pic of a half plate in my black Epics I've been taking most of the run data with:



Here's a pic of three of the many stackups I've tried. Left to right, 3K, 6K, and 12K weight twill weaves.

 

[thebtcbtq]

How long does it usually take to make a set of plates?

 

[Four by Six]

How long does it usually take to make a set of plates?

It takes me about an hour per set of plates. Actual time depends on the stackup used, resin used, and how much prep I have to do. The resin really needs to sit for 24 hours to reach full strength, and I noticed it continues to harden for up to a week.

If my additional CF supplies ever make it here, I'm ready to try making full size CF plates. I think I'm done practicing - it's time to try to make the real deal.

 

[Four by Six]

While waiting for my CF supplies to arrive, I've been wearing my various plates for casual wear, including trips to grocery and big box stores. Without the plates the Epic feels cushioned but there really isn't the "on clouds" feeling. With the plates, your feet just disappear, and it really does feel like "walking on clouds". So, these plates aren't just for making the Epic faster, they are great for casual wear. I thought maybe the casual plate would need to be thinner (less spring) than the performance plate (more spring), but it appears the same thickness and spring works for both.

 

[Four by Six]

I noticed that all of my plates fit my Pegasus 34 pretty well, so I tried my 10K run last night with my favorite pair of plates in my Pegasus to see if I could match any of my PRs. At first, I didn't feel any "energy return" but I also didn't feel the plate in the shoe either. After about 3K, my feet were killing me, but I wanted to know my 5K split so I kept running. At 5K I stopped and yanked them out of my shoes and hobbled home. My time was about same as without plates, if not a bit slower due to pain. It could be days before my feet recover. So maybe this proves it is not the spring of the plate that is making the shoe faster, but rather the interaction of the plate with the REACT foam. Certainly it proves these plates may only work well in the Epics.

My current shoe rotation is Pegasus, Epic, and ZF. My times are similar in all 3, so on race day I just wear whatever is up in the rotation. However, I think once I get all my plates made, my rotation will be simply 3 different colorways of Epic.

 

[Four by Six]

Since my feet were banged up from last night's run, I biked instead this morning. I use Shimano PD-A530 SPD pedals, the ones with a SPD clip on one side, and a platform on the other. This lets me clip in during longer training rides, but during Sprint Triathlons I prefer to not clip in as I don't like to do a shoe change in T2. I put on my running shoes at T1 and use the platform side of pedal. I tried the Epic REACT before on the bike and it was a bit too soft for my tastes. The arch of my foot would ache after the ride, telling me the shoe had too much flex. This morning I tried again with my favorite CF plates and it worked great. No flex, no foot pain. The REACT foam gave just a little bit of cushion and vibration absorption (kinda like the cranks and pedals they used to make with urethane bushings), but not so much to lose pedaling efficiency. I'm excited that I've found a shoe that works great for both biking and running - and since I have uncaged mine into a Moc, it slips on quick with no laces to tie at T1.

 

[Four by Six]

Half of my CF supplies have arrived. I am still playing with Carbon Fiber stack-ups, resins, and layup techniques. I'm also still taking run data with the test plates trying to determine the optimal configuration. My plates have gotten much more complex and are now taking me a full day per pair to make.

Below are 4 test plates. Left to right: Very flexible, Flexible, Stiff, and Rigid. The very flexible can be bent in a full circle and the rigid could be used as a chin-up bar. Weights per single plate (not pair) in ounces are: 0.25, 0.46, 0.60, and 1 with the stiffest being the heaviest (obviously).

 

[Four by Six]

The last of my carbon fiber supplies arrived last night and I spent all of today (had it off) making full plate pairs with various materials, stack-ups, and resins. They are currently drying in the garage. The resins have set already, but I'm not going to touch them until morning to make sure they are cured enough to handle. I hope one of these pairs is the perfect combination because I'm tired of working with carbon fiber.

 

[Four by Six]

I hope one of these pairs is the perfect combination because I'm tired of working with carbon fiber.

I didn't think it would be fair to only show my successes and ignore all my failures, so here is a picture of all the plate designs and techniques that did not work from last night's run. Working with carbon fiber is really nasty business, especially without all the special tools and tooling required. For example, I had several layers of gloving on and the CF cut through them like a hot knife through butter.

I made 10 pairs of carbon fiber plates, and 2 pairs came out usable. Here is a pic of the two pairs of good plates:



Of these two pair of useable plates, one is slightly stiffer than the other, but both are very flexible and springy. I will probably put the more flexible pair in my every day casual EDC shoes and the stiffer pair in my running shoes. Here is another view showing the curves of the plate:



Now I just need to take some run data...