Family bike. Triplet, sidecar?

[rocwandrer]

I have a recumbent tandem that I built about 10 years ago.

A photo, taken that first summer after I built it back in 2002:

A half dozen sets of tires or so later, it has around 6,000 unassisted pedaling miles on it. Now that we have a child, I want to still be able to go out riding together. We used to do 60 mile rides a few times a summer, and 10-20 mile rides a few times a week. The dog used to go with us for the shorter rides, pulling on the uphills. We haven't ridden this bike but maybe 20 miles in the last 2 years.

I want to either build a triplet recumbent tadpole trike, or convert the tandem to a delta trike, or, or, or. I'm thinking about a sidecar with a partially enclosed body now. It seems like perhaps the best versatility to construction effort ratio. If I do something that will accommodate the dog when she gets tired and the baby all the time, It might be nice to have assist on the uphills. I'm thinking enough assist to more than counter the entire extra drag of the sidecar for the whole ride, but applied in bursts, only on the steeper uphills.

There is plenty of room for a mid drive of some sort on the tandem, and an unused disc mount on the rear hub that could be pressed into use as a right side drive. There is also the possibility of driving the sidecar wheel, which I like from the standpoint of packaging the electrical stuff, and also from the standpoint of easily having a pedal-only tandem bike again when not using the sidecar.

Since the plan would be to use a lot of assist on uphills to maintain a workable speed (say 12 mph minimum on the steepest hill), I might also be able to use a slow wind geared hub motor intended for big wheels in a small wheel on the sidecar.

Thoughts? Input?

 

[Chalo]

There is no doubt that a kid trailer is the path of least resistance and expense in this case. But if you liked doing things the inexpensive, easy and obvious way, you wouldn't have a recumbent or a tandem, now would you?

I have a cargo sidecar bike. I like it for what it is-- but if you choose the sidecar route, make it a tilting sidecar for sure. Mine is fixed (actually integral to the frame), which is fine for loading and parking, but not great for riding qualities.

Powering a sidecar wheel will only do you any good when the sidecar is loaded. I have a brake on my sidecar's wheel, and it's the same deal: The brake is only of use when there is weight over it.

Also, be sure to have the sidecar wheel directly in line with the rear wheel, or tire scrub results in any turn. I set the sidecar wheel a bit forward to stabilize the load bed, but it's draggy and hard on tires that way.

Check out the Chariot Sidecarrier, a bolt-on solution for normal bikes that might offer you design ideas:

A triplet recumbent is just insane. Buy a used city bus if you want something that ungainly.

 

[rocwandrer]

Definitely doing a tilting style if I do it. It took me a while to get my head around why the wheel needs to be directly lined up with the bike rear wheel. Makes me think of castering the wheel (85 degree head angle, negative rake, damping grease, should eliminate shimmy), but clearly if nobody goes there, there must be issues with it.

I wouldn't have the sidecar mounted without at least some load in it so that's not an issue. The triplet idea would have the child's position upright. It would only be a foot or two longer than the tandem.

 

[snath]

Here's a thread on a sidecar that I built a few years ago.

http://endless-sphere.com/forums/viewtopic.php?f=6&t=17649

It worked pretty well but I would recommend a motorcycle steering damper at the flex joint.

Snath

 

[Chalo]

Definitely doing a tilting style if I do it. It took me a while to get my head around why the wheel needs to be directly lined up with the bike rear wheel.

If both rear wheel and sidecar wheel rotate on the same axis, then when the vehicle turns, they both rotate about a common center somewhere along that axis. If the sidecar wheel is set forward of the rear wheel, then there is no point of rotation that can be on the axis of both wheels. The more lightly loaded one scrubs along in a direction contrary to where it's pointed. The tighter the turn, the more pronounced the mismatch.

Makes me think of castering the wheel (85 degree head angle, negative rake, damping grease, should eliminate shimmy), but clearly if nobody goes there, there must be issues with it.

I thought about castering the side wheel of my sidecar to resolve the problem, but decided I could live with the problem for all the effort it would take to fix it. Potentially a castered wheel could flutter like a worn-out shopping cart wheel, but I don't think it would necessarily be an issue. It doesn't seem like the side effects could be any worse than what folks who use sidecars already live with.

Here's a little background on the matter from Everyday Engineering magazine's April 1919 issue.

 

[JennyB]

I too have thought about building a tilting powered sidecar, but never actually got round to doing it. If you get it right, the combination should stand upright when stopped, yet lean into corners as before. What happens is that side force on the sidecar wheel when it is moving causes the tyre to flex, thus steering the wheel out or in and allowing the cycle to tilt. A normal bike wheel should work perfectly; there is a lot less net sideways cornering stress than on a trike wheel.

Your recumbent tandem should make a ideal tug - lots of directional stability and plenty of space between fore and after attachment points without getting in the way of the pedals. XtraCycle now has a sidecar attachment that can be folded up when not in use. If you have a look at their videos it will give you a good idea of how low the hinge should be.

 

[dontsendbubbamail]

My vote is for a triplet with electric assist. My family started with a tandem and a trailer bike. Later we upgraded to a triplet. My bikes were all uprights. Don't think I have ever seen a triplet recumbent. Would love to see a build thread on a triplet recumbent.

Bubba

 

[TylerDurden]

Tandem-bakfiet?

 

[rocwandrer]

Tandem-bakfiet?

That's an idea. I like that it would put her in easy view of us; sort of the opposite of a trailer. For adapting the tandem recumbent, my legs (knees) occupy pretty much the entire space below the required angle from eye level to see to ride. Anything put in front of me that clears my knees blocks my view of the road.

 

[rocwandrer]

My vote is for a triplet with electric assist. My family started with a tandem and a trailer bike. Later we upgraded to a triplet. My bikes were all uprights. Don't think I have ever seen a triplet recumbent. Would love to see a build thread on a triplet recumbent.

Bubba

I'd love to have the time to build it... I actually already have some frame parts (like the kingpin assemblies with bearings) fabricated for a tadpole recumbent trike. Then I got a full time job, and the project died. I'm leery of starting it up again for the same reason (still have the full time + job).

 

[rocwandrer]

I too have thought about building a tilting powered sidecar, but never actually got round to doing it. If you get it right, the combination should stand upright when stopped, yet lean into corners as before. What happens is that side force on the sidecar wheel when it is moving causes the tyre to flex, thus steering the wheel out or in and allowing the cycle to tilt.

I had wondered if there is enough scrub resistance for that to work (pivoting, yet stabilizing when stopped). In 6000 miles we have had 3 crashes. One crash was a front tire blowout at 35+ mph on a steep downhill, the other two were very low speed tight turns on uneven ground where there was disagreement between stoker and captain about whether to stop that wasn't communicated verbally. Body English is sufficient for stability at very low speed, but when there are two sets of body english and the lean angle all in the mix... I wonder how quickly the stabilization effect goes away with speed? If I could bias in favor of resistance to leaning at near-stopped speeds, it would make my wife more comfortable with the idea. If I could bias in favor of no resistance to leaning at high speed, it would make me more comfortable with the idea I imagine tire pressure, pivot height, and sidecar load are primary factors....?

Your recumbent tandem should make a ideal tug - lots of directional stability and plenty of space between fore and after attachment points without getting in the way of the pedals. XtraCycle now has a sidecar attachment that can be folded up when not in use. If you have a look at their videos it will give you a good idea of how low the hinge should be.

My minimum hinge height is restricted only by ground clearance for cornering. This effectively means I could mount the hinge about 2 inches above the ground at a minimum. It can be two part coaxial hinge, and you are right, there is tons of room forward of the rear wheel before fouling the stoker's legs.

It doesn't look like xtracycle put a ton of thought into the hinge location. The downward extension is about the length of the quill style road stems they used as prototype mounts below the height of the existing (designed without a sidecar in mind) running board mounts.

I'm also thinking this could grow with our daughter. A kid's bike could be attached with a nearly-parallelogram linkage, so that it leans into turns appropriately

In fact, I have the somewhat unique opportunity to put a single pivot sidecar hinge on the opposite side of the bike from the sidecar, such that the sidecar leans the correct direction all the time without a multiple linkage setup, without losing the low speed stabilization feature of a single linkage design.

 

[rocwandrer]

Definitely doing a tilting style if I do it. It took me a while to get my head around why the wheel needs to be directly lined up with the bike rear wheel.

If both rear wheel and sidecar wheel rotate on the same axis, then when the vehicle turns, they both rotate about a common center somewhere along that axis. If the sidecar wheel is set forward of the rear wheel, then there is no point of rotation that can be on the axis of both wheels. The more lightly loaded one scrubs along in a direction contrary to where it's pointed. The tighter the turn, the more pronounced the mismatch.

Makes me think of castering the wheel (85 degree head angle, negative rake, damping grease, should eliminate shimmy), but clearly if nobody goes there, there must be issues with it.

I thought about castering the side wheel of my sidecar to resolve the problem, but decided I could live with the problem for all the effort it would take to fix it. Potentially a castered wheel could flutter like a worn-out shopping cart wheel, but I don't think it would necessarily be an issue. It doesn't seem like the side effects could be any worse than what folks who use sidecars already live with.

Here's a little background on the matter from Everyday Engineering magazine's April 1919 issue.

Missed this post... The shimmy can be avoided by using a head angle (kingpin inclination) forward, like a traditional bike, but steeper, and rake (kingpin axle offset) rearward, the opposite of a traditional bike. This configuration does not back up well, and may flutter on initial take off, but is very stable tracking once moving and resistant to high speed flutter.

That link you gave seems to describe the flxible company's 4 bar linkage sidecar. Cool stuff. here is more, with better pictures: http://thekneeslider.com/flexible-tilting-sidecars/

 

[rocwandrer]

Here's a thread on a sidecar that I built a few years ago.

http://endless-sphere.com/forums/viewtopic.php?f=6&t=17649

It worked pretty well but I would recommend a motorcycle steering damper at the flex joint.

Snath

That's cool! What was the problem you think could be fixed with a damper?

 

[JennyB]

I too have thought about building a tilting powered sidecar, but never actually got round to doing it. If you get it right, the combination should stand upright when stopped, yet lean into corners as before. What happens is that side force on the sidecar wheel when it is moving causes the tyre to flex, thus steering the wheel out or in and allowing the cycle to tilt.

I had wondered if there is enough scrub resistance for that to work (pivoting, yet stabilizing when stopped). In 6000 miles we have had 3 crashes. One crash was a front tire blowout at 35+ mph on a steep downhill, the other two were very low speed tight turns on uneven ground where there was disagreement between stoker and captain about whether to stop that wasn't communicated verbally. Body English is sufficient for stability at very low speed, but when there are two sets of body english and the lean angle all in the mix...

Been there, almost done that. In my case, braking to a trackstand and forgetting about the blind stoker on the back!

I wonder how quickly the stabilization effect goes away with speed? If I could bias in favor of resistance to leaning at near-stopped speeds, it would make my wife more comfortable with the idea. If I could bias in favor of no resistance to leaning at high speed, it would make me more comfortable with the idea I imagine tire pressure, pivot height, and sidecar load are primary factors....?

You'll have to build it to find out, but basically the lower the pivot is, the less side force it needs to induce a given amount of lean. I think stabilisation would be inversely proportional to speed: if you apply enough side force for a half-inch static deflection, then the sidecar wheel will be proportionately nearer or further away with each revolution. I think :? If I'm wrong (and even if I'm right) you could fix a hanlde to the sidecar body for your wife to grab if she thinks you are about to topple.

Your recumbent tandem should make a ideal tug - lots of directional stability and plenty of space between fore and after attachment points without getting in the way of the pedals. XtraCycle now has a sidecar attachment that can be folded up when not in use. If you have a look at their videos it will give you a good idea of how low the hinge should be.

My minimum hinge height is restricted only by ground clearance for cornering. This effectively means I could mount the hinge about 2 inches above the ground at a minimum. It can be two part coaxial hinge, and you are right, there is tons of room forward of the rear wheel before fouling the stoker's legs.

It doesn't look like xtracycle put a ton of thought into the hinge location. The downward extension is about the length of the quill style road stems they used as prototype mounts below the height of the existing (designed without a sidecar in mind) running board mounts.

Quill stems, eh? That's clever! I have a redundant Xtracycle I was thinking of turning into a Python-style low rider. Instant training wheels! 8)

In fact, I have the somewhat unique opportunity to put a single pivot sidecar hinge on the opposite side of the bike from the sidecar, such that the sidecar leans the correct direction all the time without a multiple linkage setup, without losing the low speed stabilization feature of a single linkage design.

I'm not sure about that. Possibly the best thing would be to have the hinge on the centre line of the bike, to minimise the rise and fall of the sidecar when it leans. I presume you've seen this link?
http://bikerodnkustom4.homestead.com/sidecar.html

Look at the "Silver Hauler" about halfway down to see what I mean.

 

[rocwandrer]

I was collecting my thoughts on the topic in text, and decided to post them here. Not exactly light reading, and I'll not be offended if nobody reads any of this; I'm just posting it in case anyone is still listening and might benefit from or expand my thinking on the topic.

In a single pivot design, there are a few cases to consider:

Pivot height - The lower the pivot, the less the sidecar leans in response to cornering and the less resistance to leaning the sidecar gives to the bike due to tire scrub. The higher the pivot is, the more the sidecar's inertia resists turns, and the more resistance there is to parts striking the ground.

Pivot on center- On center, the bike will theoretically behave exactly the the same in left-hand and right-hand corners, but the sidecar will tilt the wrong way when turning towards it, and not tilt enough to be beneficial when turning away from it. On center and extremely low, and it will not tilt much. I guess i just talked myself into believing that centered is better if the pivot is very low and the load is shared between the sidecar and the bike (load on the sidecar frame between the sidecar wheel and the bike centerline).

Pivot to the sidecar side of the bike - If the pivot is on the same side as the sidecar, the further it is from the bike centerline, the more it has to rise before it falls on turns away from the sidecar. In a practical sense, it really only rises at a decreasing rate on turns away from the sidecar, never falling because the bike never leans far enough for that. For turns towards the sidecar, the pivot falls. This means that to the degree that the load from the sidecar cargo is shared by the bike, the sidecar load tends to encourage leans in turns in the direction of the sidecar via falling of the pivot, and discourage it in turns the opposite way.

Pivot to the opposite side from the sidecar - Having the pivot on the opposite side of the bike from the sidecar means the sidecar leans the correct direction in turns in both directions., but resists turns towards it and encourages turns away from it, with magnitude dependent on the percentage of the load carried by the bike and the position of the pivot.

Sidecar vertical load position - Higher up gives more leverage to inertial forces.

Sidecar horizontal load position - Load closer to centered on the sidecar tire puts less load on the bike, reduces some of the input forces described above, decreasing the magnitude of the effects.

Sidecar pivot location and sidecar CG, front to back - configure the pickup points of the pivot correctly with the cg and wheel position of the sidecar, get the stiffness of the pivot mount correct, and the couple moment into the bike will be even. This means no net static vertical load on the pivot. This is important because otherwise load in front of the wheel and a non-centered pivot can lead to a constant tendency to want to steer one way or the other, with variations in this steering force in response to bump loads (read: unstable). I suppose this effect might be exploited to tune for the crown of the road with a pivoting sidecar.

Sidecar tire driving force vs wheelbase and offset - The further the sidecar tire is spaced from the bike centerline, the more thrust the sidecar drive has, and the shorter the wheelbase, the more the bike wants to turn away from the sidecar under acceleration.

I like to look at sample cases to figure out how things work:

- say the sidecar is 20" wide
- sidecar pivot is 12" above ground
- sidecar pivot 12" to the left of the bike
- sidecar tire on center with the sidecar load
- Sidecar tire 12" to the right of the bike
- bike HCG 28" off the ground
- bike total laden weight of about 450 lb
- sidecar total laden weight of about 80 lb
- sidecar HCG 12" above ground
- grip coefficient of 0.8 between all tires and ground

The sidecar initially leans into the turn at half the angle that the bike does, both directions. Leaning in is more comfortable for the sidecar passenger(s), even though only leaning half as much as ideal, this is still an improvement. The the sidecar inertia resists turning equally in both directions. With the pivot 12" above ground, the inertial force resisting cornering in the above scenario is the exact equivalent of having 17 lb of load refusing to lean into the corner at all. That is on par with a very small child leaning the wrong way as far as possible in every corner, but doing so in a consistent, gradual, perfectly coordinated way. Seems barely noticeable. This configuration would also allow the cg of the bike to lean up to 3.5 degrees (1.7") when stopped before the bike starts to tip. Seems very stable.

The pivot could be lower, but then it would have to be closer to the centerline of the bike, negating some of the leaning benefit. At 6" offset and height, that would be 1/4 the bike lean angle in the sidecar, and the equivalent of 8.5 lb of load refusing the lean into a turn, and still more than an inch of lean resistance when stopped.

I could use a parallelogram pivot, but then I would lose the safety blanket feeling of having the sidecar coupling resisting tilt at low speed.

I could use a single pivot sidecar, on center and low coupled, but a reaction link that tilts the child seat nearly 1:1 with the bike. Now that would be cool...and probably too complicated to actually do.