Monday, December 19, 2016

Greetings, Updates and Plans from Dan the Blogger


As some long time readers or diggers into old posts may know, I have, for quite a while now, been dealing with some issues that have been competing for the attention previously devoted to this blog. In particular, I wrote several years ago that I was caring for four elderly family members. That number is down to two, and I have finally simplified their living arrangements (sold two houses) to where they can still manage to live semi-independently. This has given me a large part of this last summer and fall to catch up on pursuing my long-term goal of developing my humble homestead on some land I own in New Hampshire. This is a monumental and urgent task for a man of my age, (62) especially since some amount of sweat equity therein has always been an important (and long postponed) component of my retirement planning.  At 85 and 87, my remaining relatives will surely and increasingly pull me away from these goals again in the near future, and so I must work around the clock while I can. My current lodging is completely off-the-grid, in a remote streamside cabin (not pictured) which I have managed to make quite comfortable except for the winter months, when provisions would have to be brought in by snowmobile. I love this Thoreau-style life, but there is no internet and even phone signals are very spotty. My projects there are varied and challenging, both mentally and physically, and I have found it difficult to switch my focus away from matters at hand even when I get to more connected places, such as the town library. (the nearest McDonalds is 25 miles away!) I do fly back to Houston frequently to help out and try to get caught up when in the city, although I am busy here as well. So I guess what I am saying is please bear with me, and don’t hesitate to comment, even if I take a very long time to reply! BTW, this problem is years old. Way back in post 106  you can see the experimental bent-roof structure pictured above when it was just a skeleton of 1x3 boards, and note that even back then I was making excuses for being out so of touch!

Recent long, cold nights, however, have afforded some time to sit by my wood stove and do some design work on the “Mama Bear” bogie design. What I am working on, specifically, is a design that is fully buildable for prototype and testing purposes. To this end I am refining the design from a standpoint of what is most practical in terms of milling, welding and assembly procedures. 

Pictured above is an exploded view of a SMART (Suspended Multi-axis Automated Rail Transport) PRT bogie of the mid-size variety. (Mama Bear) If you are completely new to the site and the concept, this is the motorized part of a hanging Personal Rapid Transit vehicle that rides within the track, so that the carriage may hang from it. This design features an unmatched combination of tighter turning, steeper climbing, faster speeds, and other many other attributes that long-time readers will recognize from earlier iterations. If it seems like there are an awful lot of parts, I would point out that many are identical, may be machined with the same set-up, can be ordered pre-cut to a shape from a steel or aluminum supplier, and so forth. A mass produced equivalent would be quite different, ideally with self-turning wheels (axial flux, hub motor driven) mounted on a stamped metal frame. This design relies on a lot of supplier-cut sheet and plate shapes, with lathe-cut “stand-offs” and lots of threaded rod, nuts and bolts to hold it all together. Below are the assembly stages. Note: I did not include every single nut and bolt, nor is this CAD file scaled to the precision customary for shop drawings. Some plate thicknesses might vary, depending on choosing aluminum or steel. If plates appear thick, they are probably aluminum.

The design starts with 1.25” square and rectangular tubing, flame-cut 1/8sheet steel and some round tube and bar, all welded.


The stand-offs are pipe which is filled with foam that is drilled to help center the pieces during assembly. 
The track switching components are assembled separately. These utilize pre-milled (for exact thickness) aluminum plate material, which is an off-the-shelf item. Aluminum can “smear” into ripples instead of wearing smoothly, so for more than a few thousand switching cycles nylon or bronze bearing surfaces should be inset for those dark blue (steel) switcher arms.
This assembled unit shows how both sets of switcher arms are sandwiched between milled aluminum plates, and how this assembly must be stout enough to never become bent to where the arms might become bound. Note that each arm’s pivot point is across the unit while it is raised and lowered by a cam near its midpoint. This geometry transfers load directly to the frame without stressing or inducing rotation to the camshaft once set into the up or down position.  

Above we see how the nearly completed switcher assembly is inserted and bolted into the frame, and additional motor mounting parts (plates and stand-offs) are added.
Finally, standard trailer wheel spindles are added as are the guide wheels and axles.

There. Now you just need 4 of Golden Motor’s 5 kw BLDC motors, ($376. Each) for a total of 26 HP, some timing belts and pulleys, 4 heavy-duty trailer tires and hubs and you are practically “good to go.” 


Lastly, sprockets are attached to the belt pulleys, (so they rotate as one) and a free spinning wheel tops off each wheel axle assembly. The sprockets act as cogs for climbing a “ladder” of dowl pins or even a mounted chain. The free-spinning wheels are used to press the sprocket into engagement, so it cannot skip. When the sprockets are engaged the tires free-spin. In this mode the bogie is extremely manuverable and can even climb straight up. That is it; except, perhaps, for adding some safety sensors to keep an eye on things out in the field, wire harnesses and other small stuff.

And one last thing – ideally the bogie should have a controller package that can make the motors behave like servos – that is to be able to numerically select (via a real-time program) the speed and number of revolutions of each motor so that, for example, the outside wheels can rotate proportionally faster than the inside ones while turning, or the front and rear wheels can rotate at different speeds when switching from cruising to climbing mode.  Post 131 explains climbing and 162 discusses using digital rotational control in general. It is, however, possible to use 4 of Golden Motor’s ($585.00 - Ouch!) controllers, unmodified, in a pinch. This is an inelegant control solution because it is inaccurate and overly complex, but it would do for now… I will discuss this option in a future post.

So, do we have any volunteers to go out and build this puppy? Oh yeah… Track will be (I would guess) about $200.00 per ft. erected… What’s that, you say? Too expensive?...  Baby bear?... Well, that is still on the drawing board.

Anyway, have a great time these holidays! Best Wishes from DTB! (Dan the Blogger)

15 comments:

James Joyce said...

Thanks for the update Dan.

No chance of building the baby bear out of Lego robotics components and putting it on a cheap wooden track then?

http://www.creative-robotics.com/About-HUBee-Wheels
http://www.creative-robotics.com/About-HUBee-Wheels?q=hubee-resources

Dan said...



I think it is important to get the subsets right here, James. The broadest subset is SMART, which is the general open source 3D methodology, (electric, in vehicle switching, transmissionless cog climbing mode, etc. SMART PRT is a subset of SMART, which narrows the field to human travel by vehicles or a size more or less consistent with automobiles. Mama, Papa and Baby Bear is a still further subset of SMART PRT combined with ordinary GRT as this refers to a scheme whereby various track sizes and vehicle weight categories may comingle and operate cooperatively. I have yet to design a Baby Bear vehicle, which must be done before finalizing any track design, and matters like splitting off or merging a Baby Bear track with a track that can carry both Mama and Baby are not trivial design challenges.

Anything that could be done with Legos would certainly not have the detail or scale accuracy to differentiate between SMART, SMART PRT or certainly not between the various “bears,” even if Baby Bear was a finished design. Sorry to nit-pic you on this little detail, but when it comes to engineering, being precise is very important, since PRT has had a long history of proposed systems that, upon close inspection, ignore physics or economic reality in some respect. I do not want to create the impression that Baby Bear is synonymous with 1-2 person PRT, when it is something quite different in that it is part of a larger framework.

That said, I have certainly daydreamed about what could be done with models. Imagine a dozen little vehicles, (even small, crude ones) that could run around a 3D track layout, climbing, avoiding collisions, stopping and going from off-line “stations” and so forth. It would be a toy rollercoaster/train set. It would be way-cool to watch, and such a demonstration could go on tour, stopping at places like MIT or Google for a few weeks, running its subroutines in the cafeteria for the amusement of diners, while allowing students and engineers time to play with the code to see who could make the vehicles do the most interesting things. There could even be organized races that simulate real world conditions that require sophisticated traffic management. Google vs. IBM! Dartmouth vs. Harvard! Such a demo would also stimulate the imagination in terms of warehousing, package sorting, store restocking and other things that SMART could be good for besides PRT. After all, how better to demonstrate the capability of delivering people than by having people see the idea at work in their everyday lives!

James Joyce said...

"matters like splitting off or merging a Baby Bear track with a track that can carry both Mama and Baby are not trivial design challenges"

Gosh, I'd never even thought about that problem. How would you split a baby bear track off a mama bear track without blocking the guide ways for the mama bear bogie wheels?

http://4.bp.blogspot.com/-h3LwIBTYL24/VLrFsXFa1TI/AAAAAAAABL8/mo35YU4GkHk/s1600/Babyw2Mama.png

Dan said...

One solution is for the track to split as Mama (or Papa) and then transition later to Baby. A Mama Bear vehicle would know not to take this turn, as it is would not even be in its map of choices. This, however, should not be necessary. Splitting off removes track running surfaces rather than adding any obstruction; Baby will have to hang on to the outer half of the diverging track while a Mama will be locked to the opposite outer half to keep going straight. Both suffer some loss of support as they cross the center open area of the track that they are not on. In the case of Mama, as shown, the wheel is wider than that void so it never actually loses all support. The Baby, leaving a Mama track, however, would probably need to be able to run in complete “half-track” mode, where it is 100% supported by the continuous, outside half of the divergent track. I show this general concept back in post 141, erroneously called “Got It!” which was before I invented the two central counter-rotating guide wheels, and before I decided not to interface with the “roof” of the track, so as to allow a PRT vehicle to run in a GRT track. (Mama Bear in a Papa Bear track) “Got it!” indeed!.. Well, it seemed that way at the time. Sheesh!

The harder problems involve compatibility between the whole spectrum of possible sizes (including possible mini-babies!) while still giving them all maximum climbing, tight cornering and speed capabilities, giving them a common location for “third rail” electric feed and similar issues. If Baby is to travel in Mama Bear track, then it is important to design Baby Bear and its track so that it doesn’t require too many additions to that Mama Bear track. It would be silly to add such cost to a Mama Bear system that may never see a Baby Bear vehicle anyway, and it would be silly to design a Baby Bear system that adds complicated or expensive aspects to the track’s architecture throughout for interfacing with larger systems that may never be added. Unfortunately these tradeoffs can only be weighed by creating and studying various alternative 3D models, which is very time consuming. Hope I don’t die before I finish!

James Joyce said...

I thought baby bear was for carrying packages, and mama bear was for carrying the 2 person 'pods' with people in them?

It would be interesting to see your design for a baby bear bogie, although now I am not sure it would advance things much, unless it turns out to be a lot cheaper to build a demo unit on a short track.

One question - ok you've told me above that building a bogie cheap buy not building it to spec would be a waste of time. But what about the track? You say steel is $600 per meter. But for a basic demo track not used very much (and with only maniquines in the 'pod') couldn't you use something cheaper such as timber?

If these layperson questions are getting on your nerves, let me know!

Dan said...

Not at all, James. The fact is that you have prompted me to review, in the last week, where I left off and to push a little bit forward. The whole division of where Mama leaves off and Baby begins is a study in compromise, and my original post reflected only a preliminary concept, which can be firmed up a bit because of progress on the Mama Bear design. It now looks like Baby Bear is going to be a bit larger than I originally envisioned, but only a detailed virtual “build” of Baby will tell, a project which is on my desk.

I ran a few dozen characters over the word limit for posting comments, so what follows is a continuation…

Dan said...

Let’s unpack some of the logic here. The “Mama Bear” is the all-purpose, 2-4 person vehicle. That is to say it can fit a person in a wheelchair and an aid, Two 100 kilogram people with luggage, both parents and two young children, (using a fold-down or popup “seat”) two adults with bicycles, etc. Being ADA (Americans with Disabilities Act) compliant makes the vehicle susceptible to over-packing since there is empty space available, but Mama Bear embraces this reality, as the greater power and size lends itself to more comfortable, longer trips at higher speeds and the kind of general performance that people expect from their cars.
The baby is a big one-passenger vehicle. That is to say two slim people could probably squeeze in, while someone over 100kg might feel a bit confined. Nevertheless most people would find it plenty roomy for getting to work or for getting around town. Baby Bear is the economy version as well as the smaller one, giving support to Bill James’ (JPods) declaration,“It costs less to move less.” Baby Bear is not ADA compliant and so would require a (road based) shuttle to be provided on request in areas without Mama Bear service. Baby Bear is more about cost than performance or comfort. The vehicles are slower, so they don’t throw as much weight around corners, and can be programmed to keep more space between them. They probably would be less cushioned, not quite as quiet, etc., as they are designed for shorter distances. The track would be supported more frequently (long spans could be engineered differently, such as adding cable-stays) Support poles could even be removable, anchored in semi-permanent modular concrete “feet”.. Baby bear is all about increasing the reach of the system on a budget – that last mile - or as a stand-alone system for close-in neighborhoods. Baby Bear would not go on Papa Bear track and would normally only share low-speed segments with Mama. (except, perhaps, for fast light freight vehicles or at night)
All track sizes can carry freight. Approx. limits for passenger weights would be up to 120 kg for baby bear, up to 250 for Mama, and over that (Papa) the track is simply stiffened with more girth, supported more frequently, etc. as needed for loads expected to go up to, perhaps, 1200 kg. At a certain point stuff should just stay on the ground! Barebones freight vehicles could carry more payload than passenger vehicles.
That price guestimate was for Standard Mama Bear track, capable of unsupported spans of 25m with multiple loaded vehicles on that span. (uses 120 kg steel per meter) Shorter spans with less load could be supported with much thinner steel, which would cost a lot less. The Baby Bear would probably be the way to go for a demo, however, being about as light and cheap as the system can be built without needing custom miniaturized parts.
I think you underestimate the scope of the task in regards to a mechanical demo, though. It would take a couple thousand man hours and tens of thousands of dollars, at least. It would need a place to be built and would take heavy equipment to assemble, etc. Although it would be like testing a car in a circular driveway, one thing is in its favor. Since these designs allow for turning on a dime a very serpentine test loop could also be very small, and turning up the speed would really put that G force cancelling design to the test! Still, it would be very difficult and expensive to ever move, so the demo would be primarily for sharing via video, and that raises the question of whether it wouldn’t just be better to do a decent simulation instead.

James Joyce said...

How much would a 'proper' simulation cost? It would probably be even more costly, but it would be nice to see some of this simulation overlaid on a real city model, like a transport corridor in London.

You'd probably have to simulate crowds and pedestrian traffic also. Otherwise the simulation could be dismissed as not being able to handle rush hour.

Also I saw some articles on hyperloop this week, and started thinking about why the SkyTran people don't promote their system as an inter city solution (apparently it can travel at 400 km/hr).

But that got me thinking. If you had hanging pods and arms that could switch from your wheeled bogies to SkyTran maglev bogies, would it be useful having a couple of extra Skytran 'rails' on straighter tracks across large cities, to act like a 'pod hyperspace' like motorways do for cars at present. This could enable larger urban areas in the future.

Dan said...

James, I’m afraid I have no idea what you mean by “proper” or “simulation,” but I suspect it would vary by hundreds of percent, depending on what you mean and who would do it. I also would be guessing. An awful lot of the cost of one time projects of any type depends on who is doing it, where, and what resources are immediately and conveniently available to them. There are simulation programs that let you draw track on a map and add vehicles, stations and so forth and plug in different speed limits, etc. I think Beamways offers one.
Skytran is really designed for speed more than for neighborhood use. It’s like strapping rockets on a taxi. When below its lift-off speed it runs on wheels just like everybody else. Frankly, I have some doubts about their business model. As speed is added wind resistance is squared, so as you go faster your trip becomes exponentially more energy intensive. This means at higher speeds, where you really have to push hard to go even a bit faster, it proportionately more wasteful to use a private vehicle instead of traveling in groups, where a long, multi-seat vehicle can have only a little more wind resistance per passenger. This all means that, in a world where energy efficiency and costs matter, the practical speed limit on their PRT vehicles is far below the theoretical one. Determining the speed range that is their “sweet spot’ vs. wheeled vehicles is a subject that is far beyond the scope of this comments section, but I believe that it is important for people to realize that maglev is not a panacea that makes going fast as efficient as going slow. At very high speeds pushing against air is like pushing against water, making it, by far, the main resistance, while the difference between wheels and levitation becomes comparatively unimportant. I have written about Skytran and their Inductrack technology back in post 143.
Lastly, the upper speed limits of Papa Bear track is a major consideration, as it is designed for group transport, longer distances, and higher speeds. It is possible that some type of track-based linear propulsion system could be used to boost the speeds of vehicles that rarely use Papa Bear track and so would normally not carry large motors. At certain speeds lift can come just from the shape of a vehicle and so track-based magnetic field generation could also further lighten and center a vehicle as well, taking stress (wear and tear) off the wheels. Again, though, going much above automotive speeds quickly becomes overwhelmingly energy intensive because of air resistance. (hence the logic of hyperloop) With a 6-10 person GRT vehicle the sweet spot can be in a higher speed range than for PRT, but there is still a limit. Ironically, bigger vehicles imply fewer trips, which implies more wait time, which requires more speed to compensate – the same old “catch 22” dilemma.

James Joyce said...

Sorry, I meant to write 'decent simulation' which were your words refering to whatever kind of computer simulation you had in mind.

I imagine there is engine simualtion software out there, which could be used to show the bogie and hanging pod actually working.

The other thing you' have to simulate might be crowds and parking in a realistic city model. I don't know if software yet exists for that?

James Joyce said...

May not use the plans of a real city, but use Los Santos out of GTA V. You'd have to populate the city with a realistic amount of pedestrians though (which it may not be able to handle). I have no real idea how to do any of that.

Rip Rook said...
This comment has been removed by a blog administrator.
Walter Harrist said...

Gorgeous design, Dan. Been a fan watching this develop for some years. You have the boggy now. I am interested to see alternative track designs.

I agree with your skepticism of the Skytran business model and the expense of the maglev. KISS (keep it simple stupid). I think it worth noting as well that at peak travel times, you don't have to be a speed demon; if you travel at a constant speed - you can compete. It takes me an hour and 20 minutes to go 20 miles on my commute; about 15 mph. Me and 10's of thousands of others, just here in Austin. Twice a day. What a waste!

I think HyperLoop is trying to solve the wrong problem. We have an inner city problem, jets are going to be hard to compete with for long distance and do a good job at what they do. On the other hand, stop-and-go traffic and bottlenecks waste huge amounts of time and fuel - and this makes it the top priority to fix.

Another side benefit note: Get 2-5-10% of people using a system - and you have taken 10% of people off the road, reducing demand/pressure for building more highway lanes and for far cheaper. The tipping point will come where people want to invest in the PRT network instead of more lanes. Network benefits multiple. Real world packet delivery, exponential growth of the network. Mirrors the growth of the internet. More people on it - more places it can go - win/win = growth.

Electric cars do not solve this. Automated cars do not solve this. PRT will change how we design and live in cities.

Glad to have you back in the saddle, Dan. Keep blogging, love it.



PS:
A statewide study by the Texas Department of Transportation (TxDOT) found that the stretch of I-35 between Highway 71 and US 183 is the fourth-most congested stretch of road in the state. It is not uncommon for motorists on this road to spend a significant portion of their commute stuck in heavy traffic.

On that patch of bad road, TxDOT estimates that travel times can be increased by as much as 73 percent, costing drivers over $100 million annually. And that's just one stretch of road-nine other Austin roads made TxDOT's list of the 100 most congested in the state.

Dan said...

Thanks for your kind words, Walter. I know it has been a while between posts, but let it be known that I am still designing, at least a few hours a week, and usually quite a bit more. For what it is worth I have discovered that those blue steering guide arms hit the track’s rack (used for “cog” style climbing) so I have redesigned them and a few other pieces. Details to follow shortly.

About KISS… Unfortunately, optimal functionality rarely coincides with simplicity. Just look at a modern automatic transmission! Our task then, is to anticipate this from the beginning and never add a single detail that might hold back such exquisitely capable engineering. Hopefully the same track will be just as useful many decades in the future, and populated with vehicles that put anything we could build now to shame.

I looked up that stretch of I-35, and noted that it is one of the few ways across the river. It illustrates one common problem with city growth in general – They plan for road widening but not for completely new routes. They could build a new bridge but all of the land on each side is already spoken for. This is where elevated microrail type systems (PRT, small GRT) could really make a difference. Both the bridge and right-of-way costs would only be a tiny fraction of most alternatives.

One issue with your sad dilemma is that with any congested artery there are relatively few drivers going straight through, but many more just getting on for a couple of miles with their entry and exit points being staggered all along the length of the route. Jumping on the freeway for a few miles is often the lessor of two evils and were that highway widened, it would quickly fill up with still other commuters who would find it newly advantageous to do the same. This, then, increases the street traffic near the highway ramps, which then forces nearby traffic to still other alternatives… You almost need a super computer to model it. In Houston we build 10 lane freeways that you can’t even get near. Crawling along on the freeway leads to road-rage. Waiting for a light to cycle 10 times before you can even reach the feeder is enough to turn you into an insane PRT designer!

James Joyce said...

Ok I'm still trying to imagine a way to sell this idea. I saw another video somewhere (that I can no longer find) where some guy buy a full scale mock up ride on top of the rails PRT pod and about 15 feet of track... and it looks like the public didn't really notice. And that probably cost the guy a tonne of hours.

So I don't believe a hanging pod on a short track would garner much support.

There are also loads of fairly cheap metrino youtube videos out there. And no real attention. So even an accurate simulated bogie and pod model might not get much attention either.

However the Cities Skylines videogame is adding a mod for mass transit systems (although sadly not a PRT):

http://www.pcgamer.com/cities-skylines-mass-transit-dlc-will-deepen-the-sim-but-more-importantly-blimps/

And some people are creating reproductions of real cities in the game (although I'm not sure of the actual accuracy):

https://www.theguardian.com/cities/gallery/2015/jul/15/bulldoze-white-house-real-cities-cities-skylines-in-pictures

There are people who work almost full time on creating assets for this game (usually buildings) for payment. Maybe someone could take a re-created real life city, pay a modder to convert the gondola system into a PRt system, and then show how a PRt could improve traffic in a city vs other mass transit options?

http://www.pcgamer.com/how-cities-skylines-is-being-used-to-build-a-real-life-city-district/