Wood Is Good: Twmpa Cycles GR1 Gravel Bike Review

Steel may be real but, for Petor Georgallou, that cliché begs the question: is wood good? Ever the curious reviewer, Petor has long wanted to test the ride quality of a wooden frame and the opportunity to ride a Twmpa Cycles GR1 gravel bike finally presented itself a few months back. But before Petor dips into his ride impressions, he shares an exchange with another wooden builder, Mark Kelly, about the characteristics that make wood an especially compelling frame-building material. After checking out Petor’s review of the GR1, drop into the comments and let us know…would you? Wood you!?

Wood Is Good….

Like most people of my age, I grew up riding steel bikes. However, it wasn’t until years later—after studying at The Bicycle Academy and repairing and building probably hundreds of steel frames—that I began to scratch the surface of what all steel can mean as a general term. I just hadn’t really understood how the material and its subtleties, and the behaviors of different alloys in different conditions, worked in different ways.

It wasn’t until after I’d had some long conversations with a few designers at Columbus that I really began to appreciate the work that goes into the sort of tube that makes a frame builder’s job easy, and a rider’s time enjoyable. Because to call something “steel” is like saying plastic, or paper. I’d be just as disinclined to try and paint watercolors on toilet paper as I would be to wipe my butt with an acid-free archival 640GSM textured Fabriano watercolor paper. That’s how different a mild steel tube is from, say, a generic Taiwanese butted 4130 tube, or how different t45 is from Columbus Omnichrom. The further you go down the rabbit hole of what steel is and why some of its variants are exceptional materials from which to construct a bicycle frame, the harder it is to make sense of sayings like, “steel is real.” Would steel be real if it didn’t rhyme? As a phrase, it serves as a weird signifier of pseudo-knowledge that lurks in independent bike shops and on stickers; I suppose it’s a better sentiment to live by than not knowing steel is good, but it conveys less than a genuine sympathy for highly-engineered, butted super-steel bike tubes. Still, it’s a starting point. So, if steel is real then… wood is good?

Weird Flex but Ok…

There are few things more frustrating than being marketed to on the ambiguous, and probably fictional, basis of a product being “green.” Wood for sure has less of an environmental impact than carbon fiber, and probably less embedded energy in its manufacture than other materials, but a frame is different than a bike, which will be built with the same components as any other bike. A huge part of the true environmental cost of industry is industry.

On a purely “green” basis, super high-grade used bikes exist, and their environmental impact is 100% lower than making anything new. I’m over the oil industry’s blame-shifting, personal carbon footprint-shaming propaganda, and increasingly people are savvy to greenwashing. For the genuinely concerned, used bikes are the “greenest,” but for realists, the individual impact on the planet of buying any high-quality bike, made from any material, is extremely low compared to a lot of other choices that someone who can afford a new fancy bike could make. Worst of all, marketing wooden bikes as a “sustainable” choice completely ignores the fact that—as the title clearly states—wood is good and that in some cases, at a nuanced level, it is even an exceptional material for a bicycle frame.

While much of the press around wooden bikes makes a fuss of their sustainability credentials, I can’t definitively say if it is a better alternative from an environmental standpoint. It’s a question worth entertaining but a more compelling question is: how does it perform, from a ride quality perspective, when it is used to make bikes?

To get a better idea of why wood might be a suitable material for a bicycle, I interviewed Australian builder Mark Kelly, who made wooden frames under the Lyrebird Cycles banner for several years. Mark is the only builder I’m aware of that has made bikes from laminated tonewood, with internal machined stainless lugs and external filets from wrapped boron fiber. His work feels more like living research than framebuilding, where his metric for success is almost solely based on perceived ride quality. Seeing as Mark is based in Australia, he couldn’t be further away from me and the following is from a back-and-forth via email in 2019:

Mark: “I am driven by curiosity before anything else. [Building with wood] came from me asking myself what made ‘ride quality.’ Of course, this implies that ride quality is a thing and that it is affected by frame materials and construction. The first experiment I did was to take my straight-gauge titanium 3/2.5 frame and selectively ‘butt’ the main tubes in a pattern that I made up based on the old Columbus SLX rifling, using an ultrasonic thickness gauge to guide the material removal. I put the bike back together exactly the way it was before and took it for a ride. I hadn’t got to the end of the street before it became evident that ride quality is indeed a thing and it is affected by frame construction. The tonewood idea came from talking to my father-in-law, who is a luthier. That led to another round of experimentation. I am still experimenting and learning and every frame is a bit different from the last as I incorporate what I learn.”

“I’ll preface this by saying that my first couple of ‘high-end’ bikes were steel and I rode them for over thirty years. Steel used to be the best material known for bikes. There are some who say it still is….I am not one of them. I think carbon is indeed a better material for the highest-end performance bikes but my standing joke is that if you are paying for a high-end carbon race bike you’ve got it wrong: if you were good enough to need that performance they’d be paying you.”

“Jokes aside, I think the race thing really distorts the bike world: A bike that is ‘better’ for a World Tour racer is not necessarily better for us mere humans. People who come to me are looking for different things, things that increase their enjoyment of riding (and coincidentally tend to increase the time they spend riding).”

Mark went on to describe how he uses a materials performance matrix (similar to The Cambridge Materials Selector) to decide which materials are the most suitable from bicycle frames, if ride quality is, indeed, the most important factor. I then wanted to do a bit of thought experiment and asked him if one could design a carbon-fiber frame to ride like a steel frame, or an aluminum frame to ride like a carbon fiber frame, etc. Aside from weight, would a normal cyclist be able to feel a difference in a blind test?

M: “It isn’t possible to do that if the time domain is considered because material properties change in different ways, so if you matched all points at, say, static deflection the dynamic behavior would be different. ‘In the time domain’ specifically refers to whether the model of the behavior has a time-variant component. Static compliance and dynamic compliance are not necessarily the same thing. Different materials respond differently as the frequency of the stress changes. Are these things detectable by a ‘normal cyclist?’ I think so and I’ve certainly met many to whom the difference is very apparent. Conversely, there will be many to whom it is either not apparent or not important.”

The conversation then shifted to Mark providing a technical explanation of how one might quantify the traits of different materials under load:

M: “I believe, but cannot prove, that the variation in loss tangent with frequency is an important trait which characterizes the way we respond to different materials under dynamic loading. This can be, and is, measured for individual materials and the information is available. The final result can be measured using an impulse response plot, common in the audio world. Much of my theorization is a crossover from a former life designing and building audio equipment. Any material that is strained and released will give back less than 100% of the strain energy: the difference is the loss tangent, I’ll leave out the mathematical description of exactly why it’s a tangent. For many materials (e.g. most metals) the loss tangent is very small and doesn’t change much with frequency. For many other materials (e.g. most composites) the loss tangent is larger but also doesn’t change much with frequency: carbon epoxy is one such, it is an inherently damp material because it loses energy every time it is flexed and the amount of energy lost is pretty constant even when you change the frequency of the flex.”

“Some materials start out with a low loss at low frequency, but increase loss as frequency rises and this includes tonewoods. The net result is that they respond in a unique way to stress. Impulse response refers to a way of testing this: an impulse is a sudden change in momentum, in acoustics it refers to a sudden change in sound pressure. Via Fourier Transform theory, such an impulse contains all possible frequencies so the response to the impulse is a way of testing how an object responds to different frequency stresses. A poor man’s impulse response test: bang the frame with a hammer and listen to the result. You’d be surprised at what you can learn (if it is carbon and it goes crunch you just learnt that you hit it too hard).”

Think Globally, Shop Locally…

Having had this back-and-forth a few years ago, I’ve been pretty keen to test ride a wooden-framed bike for an extended review. I did my best to conceal the fact that I was doing a little happy dance when I was asked to review the Twmpa GR1. The GR1 is designed and built by mechanical engineer and furniture maker Andy Dix in Haye on Wye, Wales. There’s a relatively small number of wooden bikes on the market, but I’ve not seen one in the flesh that looks as sensible and normal as the Twmpa. So, I was extremely pleased to meet Andy halfway to decant the GR1 from his van to mine.

The Frame

The Twmpa GR1 is made from ash, which is laminated to increase strength at the apexes of the tubes at the joints, with aesthetically wavy, CNC-milled jigsaw joints. The front of the frame is made as two 2D triangles, before being CNC milled both internally and externally as two halves, which are then glued together. The stays are each made individually with the formed chainstays made from ten laminations of ash, which are then bonded to the front triangle. The front triangle is effectively continuously internally butted and hollow, which means the frame can be both incredibly strong and incredibly light. As a process, this brings out the best in the material: ash on ash with hollow sections, designed to save weight (the bare frame is 1,700 g), but also tune flex, eliminating stress risers.

The GR1 has a BSA bottom bracket, tapered external headset, 27.2 mm seatpost, and it accepts the standard thru axles that you’d expect on a modern gravel bike. There’s clearance for a 700c x 38 mm tire, which by modern-day measures is more of an all-road clearance than a grav-grav clearance. It has internal cable routing guided by a black nylon sleeve glued into the frame, which is just about visible on the surface. The bike I had on loan was an early model; updates on the latest version include more clearance for a 700c x 42c tire, a T47 BB internally reinforced with carbon fiber, as well as beefier chainstays, also reinforced with carbon fiber for added lateral stiffness. The nylon internal routing tubes are bonded differently on the most recent bikes and are therefore not visible at all. I’ve not ridden that bike, but having ridden this one I’d agree with the updates as a way to improve the ride quality. Another update is the downtube logo, which on my bike I’d call aesthetically disturbing by being too small for the tube diameter. The new logo matches the dimensions of the frame more pleasingly and is just better in every way.

The Finish

The frame is finished with five layers of epoxy that protects it from the elements and contributes to the strength of the wood, as well as a layer of 2k polyurethane to protect the epoxy coatings from UV. It’s nice if you’re into that kind of thing, which isn’t really to my taste. This is all incredibly personal but, in my experience, it impacted my enjoyment of the bike. Every time you stop, people want to talk about how the bike is made of wood. Cyclists and non-cyclists will stop and say “Your bike’s made of wood!” a statement, not a question, beyond which they have no will to genuinely engage with what it means to ride a bicycle made from wood. At best they’ll follow up with “isn’t it heavy?” or “Is that strong enough?” It didn’t quite get a tall-bike level of attention from the laypeople, but it wasn’t far off which became pretty tedious pretty quickly. If I owned a Twmpa, I’d get it hydro-dipped in Realtree to camouflage the real tree. If you’re into the wood look and like to chat it’s great. Otherwise, I’m sure paint is an option. Lastly, the machined aluminum dropouts were nicely anodized in brown to match the headset and cranks.

The Build

The bike arrived looking smart, with a SRAM Force/Eagle AXS mullet drive train, INGRID cranks, Enve fork, Chris King headset, and Thomson finishing kit with Easton bars. I fitted my own SQlabs 414 Infinergy Active, prototype saddle (which I’m a huge fan of) and a new set of HT M2 pedals which I was sent to review. The wheelset was DT Swiss PR1400 Dicut, which was fitted with a pair of handmade Challenge Gravel Grinder tubeless tires. The SRAM drivetrain did what SRAM electronic drivetrains do… it just worked. Each shift wasn’t an overwhelming wave of sensory ecstasy but it never missed, and zero maintenance was required over the review period on a bike which has been beaten about by a number of people before me. The brakes, paired with 140 mm rotors, were more than powerful enough on both road and off-road. Under hard braking, you can feel the frame flex significantly, and whilst that’s weird at first, it was fine once I realized that it was ok! There was no judder and even with the flex everything felt good and sure-footed and fine.

Over the past couple of years, I’ve been riding SQlab’s saddles, with a specific preference for the active rail variant. The rear of the saddle has an elastomer, which means that the back of the saddle can slightly rock from side to side. It is similar to having a more flexible top to the saddle, but not quite the same. I’ve noticed a tendency to just sit in one part of the saddle more rather than shifting around, which means I can pay a bit more attention to bike fit, which in turn helps me feel less uncomfortable on longer rides. That’s the measure of comfort for bicycle saddles: not sitting on it and feeling immediately comfortable, but forgetting about the initial feeling, and also not thinking about comfort at all on any length of ride.

I found the Challenge tires, well, challenging. They are essentially a dry-weather cyclocross tire in an illegal size; they’re super light, super supple, super fast-rolling, and surprisingly robust. I hated them. While the ride quality in dry conditions was superb, they were so supple and had such a specific tread profile that they were only fantastic in ideal conditions, running 33 PSI, in the dry, on smoothish dirt, or grav grav, without anything so chunky that they’d pinch. Perhaps they’d run a little better on a wider rim? I’m not sure. Too little PSI, and they roll on the rim and feel sluggish and pinch all the time. Too much PSI would make them feel harsh and boingy, like basketballs, and would have a negative impact on grip. In perfect conditions they were perfect, but in real life they were fussy. I understand them as a race tire that would be an excellent choice for a given course in given conditions (given that there are other tires to choose from for different courses and conditions). I found them distracting because they have a lot of the same super-soft ride characteristics as the frame, which made it difficult to differentiate softnesses. I switched them out for a set of Schwalbe G1Rs which are my current favorite gravel tire.

Another standout part was the HT M2 pedals. They come in two versions: steel axle and titanium axle. The pedals I received were the (hollow) steel axle version which were definitely on the lighter side, and unlike their titanium counterparts don’t have a stated rider weight limit. Overall, I’ve been super impressed with these pedals and I will be sharing a more in-depth review soon.

Not seeing the trees for the wood

The GR1 has a very particular ride quality which took me a minute to understand. It flexes a lot more than expected, but it also flexes very differently, and the sensation of flex carries a different meaning in ash. To start with, I rode it mostly on roads and hard-packed dirt paths, and in those situations, it felt great. Weird but great. It didn’t initially inspire confidence to get out of the saddle to sprint or climb, and because the build was fairly light, it felt somehow flimsy which didn’t inspire me to push it off road. There’s something in it of the sensation of riding a 1950s British lightweight with 27 x 1.25-inch tires, and 1” or 1 ⅛” Reynolds 531 tubes, except it’s somehow springier and lighter. It felt flexy but also fast, like it was begging for big rides, like a dialed rando or audax bike that really needed to stretch its legs for its value to become apparent. In the same way that a rando rig is designed for poorly paved roads rather than chunky modern gravel, I felt reluctant to take the Twmpa out on anything chunky.

For the purpose of a fair review of what felt like a refined, luxurious all-road bike, urging me to put in more miles than I normally would, I felt obliged to find its limits off-road. I called Andy to check if it was ok because I felt like it might not be. Andy assured me that the design had been ISO-tested for gravel and passed. ISO tests don’t lie, so (somewhat reluctantly) I strapped my helmet on and headed to Covert Wood for some dry, rutted trails. Trails on 38c tires just didn’t feel like a good idea, and the geometry of the GR1 not being particularly progressive left me a little apprehensive.

However, it was in the woods that the Twmpa really felt at home. Surprisingly the more I pushed it, the better and more sure-footed and comfortable it felt. By the end of the day, I was smashing it around like a maniac over roots and rocks and rutted washboard trails, all of which it sailed over with magic carpet-like floaty smoothness. Imagine having 20 mm of travel through the entire frame, but with no rebound and perfect damping, but in a way that somehow doesn’t sap your energy. That’s what a well-designed wooden frame feels like. The frequency response curve of the material absorbs all the small stuff but doesn’t touch the big long slow pedal strokes. It also doesn’t feel unpleasantly dead, you can still feel the sensation of losing traction or hitting objects, but just not in a way that beats you up over time.

For context, at the same time as beginning this review, I was also working on my review of the Brother Cycles Mr Wooden, which is itself a pretty dynamic, relatively under-built, rim-brake bike with much plumper 650b x 48c tires on wider rims. In comparison to that flexier than normal, fairly ordinary steel bike, the Twmpa felt not only much faster but also much smoother, more comfortable, and more sure-footed off road—and it had much narrower tires. Unlike the Brother, it is a serious performance bike, it offers performance in a kind of understated way. It has a uniquely luxurious, soft-but-fast ride quality that isn’t for everyone. Because of the unique material flex, it can cope with having slightly narrower tires than you’d expect on a modern gravel bike, without beating you up on trails. I’m hoping to hold onto the GR1 for a bit longer through the winter to have a go at winter riding with a narrow, but more mud-oriented, tire that cuts in for grip rather than floating.


The Twmpa Gr1 is a light, fast, performance leaning all-road bike with the unique ride quality offered by engineered ash. It’s wonderfully floaty and has a dreamy suppleness unmatched by anything I’ve ridden before. It’s not a race machine and might not be suitable for everyone’s style of riding. It rewards riders on long rides over uneven terrain, and in that way it would make an incredible rando/audax/endurance road rig. It rewards riders who value comfort, however, it’s perhaps not the best choice for sprints or competitive climbing, although it was remarkably stable climbing off-road. The Twmpa GR1 has been a hugely rewarding bike to review, and based on the time I’ve spent riding it, I would 100% explore the idea of riding wood on a long-term basis, based on its performance and inherent comfort as a material. One nitpick is that I would like the option for a wider tire for gravel but the current GR1.1. somewhat addresses this by offering an increased clearance of 700c x 42 mm.


  • Unique ride quality
  • Super comfortable
  • Loads of flex-derived traction off road
  • Sounds nice (weird but true)
  • Eats miles
  • Takes a beating like a champ
  • Does not work harden or fatigue over time and is resistant to UV
  • Make friends with laypeople
  • Nicely made with aesthetic miters


  • Can feel a bit sloppy pushing hard out of the saddle
  • Limited tire clearance
  • Ash dents more easily than other materials
  • People asking if it’s made from bamboo
  • People talk about their friend who made a bike from bamboo
  • The unsolicited mansplaining on the positive environmental impact of bamboo bikes
  • Flammable.
  • Expensive; framesets start at £3490

See more at Twmpa Cycles.

Twmpa shop photos courtesy of Jim Cossey