Geometry critique

[bushtrucker]

Yo love the idea of this thread. Figured I’d share this as I had a bike built to a similar brief in 2020 — off-road/tourer that would do double duty as my primary MTB. This was the initial drawing I sent the builder (Darren Larkin AKA Larkin Cycles, based in LA at the time). For reference I’m 183cm (6’0) with a 89cm (35") PBH — all legs and no torso.

The final design changed based on some feedback Darren had but it was similar enough in the end. A few thoughts on the bike after 2 years of riding it.

• Seat tube could be a touch steeper for a dedicated trail bike but 73° deg seems to be where I’m happiest on longer rides. I agree that the best thing to do is find a bike you like to ride seated and base the rear end off that. A little steeper will be fine but I wouldn’t go all out there. If you’re touring on it you will be sitting for most of the ride.
• The HTA is perfect (final design was updated to be 67° sagged, shown above at 67° static), I wouldn’t want the bike to be any steeper. Currently riding it with a Helm Coil, 44mm offset, 140mm travel.
• I went with Paragon Sliders and have ended up with the stays at around 450mm. I played around with them a bit when I first built the bike and can’t really notice any difference in handling with the stays 10mm shorter so went with the extra mud clearance.
• 50mm ish BB drop seems about right for a 29er, I rarely get any pedal strike and can get pretty lazy with my feeting when I’m tired.
• The bike has a very high hand position and that makes a huge difference to me for longer rides. I’m using 37mm rise bars with 15° backsweep (Hunter Smooth Moves).
• A long headtube is great. It makes for significantly larger framebag space than I’ve seen on equivalent bikes. I can fit my tent w/ poles, all tools/spares/pump and a 3L water bladder in my framebag and still have room in the front triangle for a 800mL bottle mounted to the seat tube.
• If you plan to tour on it defo add some kinda rear rack mounts. In my experience a rack is far better way to carry gear than any seatpost mounted bag. Especially if you throw a dropper into the mix.

Photo below of the bike fully loaded for a 3-day / 200ish mile ride we did early last year. 40% sealed, 40% gravel, 20% rough 4x4 roads/singletrack. Yes it’s a compromised design but I would say I’m 90% happy with where it’s at considering what I set out after. Good luck with the project and post us some updates!

[Fahzure]

I am 6’3" and have played around a bunch with long reach/long effective reach (reach +stem). 3 of my dual suspension MTBs have reaches greater than 500mm, one is 530mm, with effective reaches all around 580mm. I can’t see going much longer, as the bikes, especially the 530mm, feel noticeably sluggish and less maneuverable until they get going 15+mph. You will need a longer seattube for your height/leg length, probably north of 450mm, or you will have to get a special seatpost.

[Fahzure]

In the '90s, for private residences around Moab/Four corners.

[poopieprancer]

The reason I was thinking that was the the 746mm number I assumed was your BB to saddle height, which I thought implied that you were of average height. Is that number for something else? Anyway, is the 699 number going from the saddle position to the hand position? I’m not intimately familiar with the PVD design approach so forgive me if that’s an obvious question.

Being 6’4.5" myself, and having gone through a few iterations already for an all-round performer geo similar to @bushtrucker’s design, I’d echo what others have said about designing for your seated fit. You can buy back some reach by erring on steeper side of STA that you find comfortable for flat land pedaling and go from there.

Does anybody else pretty much ignore actual frame reach and focus on reach at saddle height? I find those XL frames with tiny head tubes very misleading in the reach department.

[bushtrucker]

Yep. I use the “handle to seat tube axis” and “handlebar drop” measurements in BikeCAD to get an understanding of where my bars will sit. That’s only based on the centre of the handlebar clamp so the actual hand position on the bar differs but I found handlebars too annoying to model in BikeCAD and all my offroad bikes have kinda similar enough bars on them anyway.

The way reach is shortened when you put tall handlebars on a slack HTA bike is pretty significant in my experience. Even more so with steep STA’s. The ST and HT start converging a fair bit. I’m only 6’0 but I can get a way riding a lot of XL bikes because of my long legs.

[Daniel_Y]

@wizman

For ref. I’m 6”3/190.5cm and weigh about 75kg

IMO, that is pretty far the average height and far below the average weight! Most of the wisdom of bike fitting and handling revolves around 5’10, so I really encourage you to find out yourself!

Note: I use unsagged geo for various reasons. To convert to sagged, steepen the STA and HTA 1 degree and 10mm to the reach

I finally took the time to write out my method for fitting mountain bikes. What you are designing is really a flat bar gravel bike. In that case, the seated fit matters the most, so I would optimize around that. I will follow up with an example.

1) Saddle Height is the Characteristic Length

I start with the saddle height. Everyone pedals slightly differently due to preference or habit. From low saddle mashers to high saddle tap dancers. The seat high encompasses:

• pedaling mechanics
• cleat position
• shoe stack
• crank length

2) Determine STA

The STA is dependent on the terrain and use, a bit of a personal preference, kinematics, and marketing.

If I were to generalize:

• 73-74 for flat (road, gravel)
• 74-74.5 for general mountain biking
• 75 for steep or techy climbs

Keep in mind:

• I use unsagged geo! One of the reasons why I do this is that when you climb, you are riding unsagged geo.
• Road and gravel bikes typically have 10-20mm of seatpost offset, which is like another 1-1.5 degrees slacker STA!

A few things to keep in perspective. Even at a really tall 800mm saddle height, 1deg of seat tube angle is ~15mm of fore-aft adjustment.

• people’s femur lengths can easily be +/-20mm
• A saddle’s “sweet spot” relative to the rails can be +/-10mm
• The saddle rails can accommodate +/-20mm of adjustment.

4) Pick the HTA

HTA is going to be terrain, personal preference, and marketing driven.
To me, the HTA and fork travel needs to make sense with the STA and the intent of the bike. If you have steep descents, you probably have steep climbs too! Paring a slack STA and slack HTA is just asking for the bike to lift off climbs.

A few more notes:

• Riders taller than 6’0 might want to consider a steeper HTA to keep the wheelbase in check and weight on the front wheel
• Shorter riders need narrower bars, which don’t give you the same leverage to manage a slack HTA. I recommend going steeper (or less trail) if possible.

3) 5% Bar Drop (saddle to bar clamp)

I aim to have the saddle within 5% of the bar drop with 20mm of spacers. This gives the user room to play around with their fit.

In this example, Saddle_Y = 771, so the target_bar_drop = 39mm. A 140mm headtube gets us there.

Notes:

• Shorter riders <5’6 will struggle to have bars lower than the saddle
• Taller riders should use high-rise bars >40mm to keep the wheelbase in check
• I don’t design to grip ends because that dimension is impossible to get from customers
• I don’t design around a particular bar, because I expect the user to switch bars
• Bar rise is used to fine-tune the fit
• I don’t like swept bars (>9deg) on our bikes
  • the most I recommend is the PNW range

4) Match Stem Length to Steering Geometry

I think its important to match the (effective) stem length to the steering geometry. I visualize effective stem length and trail as two levers that are coupled together:

The combination of the steering input (L1) and the steering response (L2) should make sense for your application:

There are four permutations of L1 and L2:

L1 (stem)	L2 (trail)	Steering Input	Steering Response	Example	Result
short	short	fast	fast	A road bike converted to swept back bars	Very twichy bike
short	long	fast	slow	Mountain bikes	Stable bike that rejects disturbances
long	short	slow	fast	Road and gravel bikes	Nimble bike with stable steering
long	long	slow	slow	???	Bike does not want to turn

There is no magic number here because there is a lot of personal preference and people can get used to anything. However, I still think there is a correct stem length and bar width for your application. If you are going with a steeper geo (67-69HTA), try designing around a longer stem (60-70mm). Not fashionable, but you will probably like it.

Side note: The effective stem length is going to depend more on your bar sweep, width, and roll than your actual stem length. So be careful! From Richard Cunningham and MTB handlebar dimensions

What about grips behind the steering axis? I personally don’t like the feeling. I feel disconnected from the front wheel like I’m throwing a wheelbarrow at rocks. As a byproduct, the negative stem pushes your front center further out by 20-30mm, which I don’t think suits many designs.

5) Rider Compartment and Reach

The butt-to-bars dimension is a bit tricky and subjective. It can be affected by:

• saddle angle and sweetspot
• bar roll
• bar sweep
• bar width
• personal preference

The best way to figure this out is to use one of your existing bikes to get a working baseline.

• does it feel cramped?
• will you use a wider or narrower bar?
• do you want to go longer or shorter?

Or, this would be a great time to look at your RAD measurement.

6) Tuning the weight balance with the chainstay length:

You guessed it, the weight balance is also subjective and affected by marketing! For the weight balance, I compare the rear-front ratio (RFR).

From this post:

Assumptions:

• Flat ground
• weight is through the bottom bracket (ie. standing up)

Rear wheel force: F1 = L2/(L1+L2) * Mg
Front wheel force: F2 = L1/(L1+L2) * Mg

The rear/front ratio is really easy to calculate because most of the terms cancel themselves out:

rear-front ratio: RFR = F1/F2 = L2/L1

	Less RFR	More RFR
Climbing Traction	Rear wheel slips	rear wheel grips
Climbing	front wheel sticks	front wheel lifts off
Desceding	front wheel hangs up	rear wheel takes the abuse
Flat Cornering	even weight distribtuion	front wheel can wash out
wheel lifts	planted	poppy

For my mountain bikes, I found RFR of > 1.88 fun, but on the limit of what I consider a well-rounded hardtail.

Keep in mind, the assumption that the weight goes through the BB is a big one. People come in all shapes and sizes and weight distributions. Riders have different styles which shift their weight around. The RFR is more of a tool to compare your preferences to you own bikes bikes rather than an absolute goal.

7) Massaging the Numbers

This is when I take a step back and look at the bigger picture. I fill in the blanks: tube spec, bb drop, tire clearance, seat tube length, aesthetics, headset standard, fork offset, etc.

I will iterate through all steps 1-6 again, and tweak the numbers.

• Did the wheelbase creep too far?
• Did I stray too far from the design intent?
• Is this too big of a change to what the customer is used to?
• How confident am I in the customer’s provided numbers?
• Maybe the customer hates their wheels lifting off, ill shorten the reach and spec a longer stem
• Maybe the customer can’t wrap their head around a 450mm chainstay
• Maybe the customer just wants something different

For an endurance mountain bike, this is probably where I would land you:

• RFR 1.80 for more balanced ride
• 60-70mm stem with 40-50mm riser to keep the wheelbase in check
• Bar drop of ~7% your saddle_y height (more than 5% because I recommend a big riser)
• The reach can be tweaked to your preferences/ what you want to try

Here it is compared to your RAD method. It’s not a direct comparison, but the truth is probably somewhere in between:

[bushtrucker]

Hey @Daniel_Y. That’s a great summary and something I’m gonna come back to. Cheers! I wish that Bike CAD had a feature to toggle between sagged and un-sagged geo.

One thing I would add is that the initial post mentions touring. When you say weight goes through the BB I think luggage need to be considered as part of the design since how you load a bike will make a difference to handling. In my experience getting as much weight as you can centrally/into a framebag will maintain a pretty neutral feel. But you’re always limited on space there. Front end weight is ok and low trail bikes with a large front load are actually pretty good for road/gravel riding but too much weight over the bars can get pretty sketchy off-road. For a MTB style design I think a rear rack is really the go. The downside is that carrying 10kg+ of gear above the rear axle is gonna throw out the rear end a bit. This is where longer chain stays can really make a difference. There’s also added benefit they provide in adding a bit of heel clearance to a pannier type setup.

The more weight you need to carry back there the longer I feel like you want to go. The extreme case of this would be something like a Big Fat Dummy where the rear RFR is like 0.82, 720mm FC, 875mm (!) RC. I maxed mine out at around 25L water + 5 days food for a 500km stretch in outback Australia with no towns. The bike was too heavy for me to lift but it still handled pretty much the same as when it was unloaded (which is kinda like a bus haha). I would love to try a bike with a RC around the 650 mark for this kind of riding.

[poopieprancer]

This one rides like a short bus… 600 RC. Never really used it for hauling much weight except when I strapped some fire wood on there. I was able to wheelie it though. RFR of 1.23 it looks like.

[Fahzure]

Daniel’s way is methodical but I think misses an early step: determine the fork travel and axle to crown. I believe that most frame designers start with an amount of fork travel, or in the case of a rigid fork, tire and fender clearance needs. With this info in hand, you can resort to a simple formulation for HTA. Most bikes with front suspension rely on a Base 71* these days and you can approximate the HTA for a given travel by subtracting 1* for every 25mm of fork travel. Daniel’s example is a 71* Base with a 5in travel fork resulting in a 66* HTA. 72* Base used to be more common and you still find it on shorter travel bikes, while 70* Base is sometimes found on longer travel (6in+bikes). Of course, these are industry conventions, but you don’t want your HTA much steeper than 72* at full bottom.

One other industry convention that Daniel notes is slack seattube angles. For road or rolling terrain gravel terrain, those numbers may not be far off, but for any type of climbing greater than a 5% grade on soft surfaces, a much steeper seattube is in order. When MTBers encounter steep sections and moves, they cannot get out of the saddle like roadies, because of rear tire spin. If you are generally MTBing a STA of 76*+ will be great, if you live in the mountains and do winch and plunge rides, you may find 80*+ is better (remember a 80* STA on an 8* climb is the same as a 72* STA on flat ground). A steep STA relaxes the core and upper body on climbs because you no longer have to pull yourself into the handlebar, it’s like riding a unicycle. I currently have an 81.5* STA and would like to try 82* for my Rocky Mountain and West Yorkshire winching. (Works well with those longer reaches, too, putting more weight on the front wheel when climbing for less wandering.)

[anon91558591]

I don’t base the head tube angle on any fixed number minus fork travel, and I’ve never heard of anyone doing it that way, but it’s an interesting idea.

Seat tube angle has more to do with physiology and preference than anything else, and it’s nice to see more diversity there, but most people aren’t riding up stuff so steep that they can’t get up out of the saddle, or are having trouble keeping the front end down. If you live in BC, sure. But there are a lot of places in the world that are not at all like BC.

-Walt

[wizman]

I thought it would be interesting to model my current bike to get a better understanding of geometry and my relationship to it…

I think the discrepancy between the ST angle of the model and the photo could be due to improper tombstoning. Regardless I think it’s interesting that the RAD model locates my hands in a pretty similar position to their actual location on the bike.

Geo was taken from the '93 Kona catalog. My bike is the hahanna. You can’t polish a turd but you can cover it in glitter!

@Daniel_Y

The best way to figure this out is to use one of your existing bikes to get a working baseline.

• does it feel cramped?
• will you use a wider or narrower bar?
• do you want to go longer or shorter?

I currently feel (and look) like i’m riding a childs bike. I’m not trying to build a hammock but i’d definitely like to feel more in the bike as opposed to on it. This bike is definitely on the short short twitchy af end of stem and trail length. I’m going to implement your suggestions into my (read your) CAD model, and do another overlay. I can’t thank you enough for taking the time to do this, i think i’m starting to wrap my head around the core concepts.

EDIT. I’m also getting the sense that it’s very difficult to model a frame without also modeling the key touchpoints e.g saddle and bars, going to look further into that

[bushtrucker]

I reckon it’s a smart idea overlaying geometry this way. You can only build from what you know but it can also be good to check your design against what others are doing.

I’ve found Bike Insights to be a good tool for this. Most mass produced models are already loaded up there and you can add you own custom geometries too. The only annoying part is manually inputting custom values but you can always edit the geo if things change later. Since you can’t imput stem/spacer data it’s mostly based around reach/stack numbers. So not as nice as knowing your exact handlebar position but a decent starting point.

Below is an example of the custom bike I posted above compared to my old MTB. Not quite the same vintage as yours @wizman haha. Pretty quickly you can see that the new bike is a little longer and a lot taller. I’m now using a lower rise version of the same bar, a few less spacers and a shorter stem. The STA remained the same but I went from a 20mm setback seatpost to an inline one. Small but significant changes.

[Daniel_Y]

@wizman

Regardless I think it’s interesting that the RAD model locates my hands in a pretty similar position to their actual location on the bike
[…]
I currently feel (and look) like i’m riding a childs bike.

This is why I think while RAD is a good framework for measuring and comparing fit, Peter and Lee’s suggestions for the angle and length are based on what they like. IMO, all bike sizes, components, and fitting convetional wisdom revolves around 5’10 people. once you stray away from that baseline, a lot of the advice does not add up.

With my method, your bike is a lot bigger than the RAD suggestions. I know my method is not perfect yet, but weve had a lot of positive comments from many customers on the full specturm sizing.

@wizman

This bike is definitely on the short short twitchy af end of stem and trail length.

Does it feel twitchy while riding? This is my main issue against bars with a lot of back sweep (>10deg). I know that they are comfortable, but IMO they compromise peoples fit and handling so much. Unless you run a massive stem, it shortens your reach, makes your steering twitchy, and takes the weight off your front wheel (which makes it even twitcher).

I am not saying a swept bar can’t work, but I think people just slap them on bikes like stickers without thinking about the consequences of the fit and handling.

@wizman

EDIT. I’m also getting the sense that it’s very difficult to model a frame without also modeling the key touchpoints e.g saddle and bars, going to look further into that

This why I use bikeCAD for fitting, fusion360 for design and fabrication.

To get accurate overlays, you need to stand REALLY far away from the bike, crouch down so you are level with the center of the frame, and zoom in

Here is a high quality diagram:

I took these photos to double check the HTA of our frames after welding. They were taken standing across the street and zoomed in with a telephoto lens. You can do the same with a cellphone camera.

[anon91558591]

What’s wrong with swept bars and a long stem? If you like the hand position that’s a great way to go.

-Walt

[Daniel_Y]

@anon91558591

What’s wrong with swept bars and a long stem? If you like the hand position that’s a great way to go.

Nothing! However, 90% of the bikes I see with swept bars are using a short <50mm stem… IMO, people gravitate to the short stems for fashion and marketing rather than fit and handling.

[liberationfab]

I’m running Tosco bars with a 130mm stem but the grip position is about even with/just behind the steerer tube. I agree that they’d muck up handling (and run into knees) if run with a short stem/short reach combo.

[manzanitacycles]

The last bikepacking bike I made uses the VO Crazy bars with the rider’s hands behind the steering axis. Guess what, it handles great.

The bike is long front and rear to add a little extra forgiveness.

Not every bike is intended for cornering at the limit and maximizing front wheel traction.

IMO people get stuck on geometry and fit absolutes like:

• You can’t ride with a front load unless you have low trail
• If your hands are behind the steering axis it’s gonna suck
• Take your height and divide by X to get your reach or whatever the latest fit guru is saying

It’s pretty hard to make a bike that isn’t fun to ride. I still think @wizman should make something a little goofy just to see how it rides. Put sliding dropouts on it to see if you notice a difference with different chainstay lengths. Get an angle adjusting headset. Try different crank lengths. Go wild. It’s the perfect time to experiment.

[bushtrucker]

I’m currently running the same 15 deg bars on a both a 35mm stem (67 HTA) and a 80mm stem (73 HTA). Each bike feels right and the different stem length is really just accounting for different reach/eff tt on each frame. I don’t really think there’s any rule of thumb as far as bar shape and stem length goes. Whatever gets the hands where they want to be!

Edit: That was supposed to be a reply to @Daniel_Y.

[bushtrucker]

Love the stance that bike. With the long RC it’s so balanced. Great choice of colour too!

[Daniel_Y]

Re: effective stem length

I think all the parameters of bike design fall on a spectrum. The relationship between steering input (bars and stem) and steering response (trail) is a broadband spectrum, meaning a lot of combinations work.

That being said, to me, there is still an optimal relationship somewhere in that spectrum, and I’ve done a bunch of modeling and testing with myself and others to get a sense of where on the spectrum my intuition is. Everyone has their own experience and ideas that shift this spectrum around.

Zooming out a bit, and putting this discussion into context: if I were to score the importance of the parameters of bike design:

• Fit = 60%
• Weight distribution = 30%
• Steering Characteristics = 10%

All three of those are coupled together. By the time you figured out your fit and weight distribution, the stem length and handling characteristics probably already figured themselves out.

To me, it’s not a black-and-white: grips infront good, grips behind bad. However, when I see grips behind the steering axis, it’s usually a flag that there is some compromise to the fit, weight distribution, and steering geometry, especially since the original poster @wizman is 6’3! But again, bike design is a broadband spectrum with a lot of subjectivity. I’m just trying to provide a framework to guide people to their own conclusions.