Building a Round Door - PTJ 2024 Project, led by Sky Huddleston of Liberty Rocket Stoves!

Jerry McIntire wrote: better yet, add an overhang to protect the door from rain and sun.

I do this for the bottom of my wooden boat, it stands up to grinding on a beach.

Nice shape!

Adding a front porch roof to the Hobbit Home, so the round door there will be sheltered from the elements.

The epoxy i used was from my last canoe build... wanted to use it up - still had some left over after building the bench for my house...

As for the shape of the door, i agree !  Not truely a round door, but 50 feet from the road, it sure looks like one. 😁

Darn thing will not let me post pictures again... will try later...

At the SKIP event we were talking about the plate steel and Kyle came up with an alternate core idea for the door.  Paul wasn't opposed to it and suggested I mention it here for the group's consideration.

Take some 1 or 2" tube steel.  Make an octogon from it (or pentogon or dodecahegon) with some cross bracing spokes.  That should be pretty damn stiff if it's welded by someone capable.  Two of the tubes could extend out of the door to be part of the hinge (if desired).  Otherwise the tube steel core would just be sandwiched between the oak on the faces.  Insulation could fit inside the tube frame and be added on outside it as well with some detail work.  

The tubes would also be easier to attach to than a 3/16" thick plate and be more available at steel shops.

Here's a crude sketch.  My drawing tool only goes up to hexagon but pretend the tube is at least an octogon.  The dark red parts are the hinge barrel.  Once again, having the hinge be a part of this core is just an additional idea.

I realize this may not be the heavy duty, solid wooden structure, that the majority of posts on here are geared toward, but i just gotta say ....

My front wall panels have been ordered !!

I have been worrying about this for the last three months, as i was finding it very difficult to get this item.  Not only is it the first thing people will see ( exterior front of house) but i also wanted the clean white look of steel siding , with a healthy dose of insulation.

These will do the job... 10 inches thick, rated R42, no thermal bridging.  When i described how i wanted to cut a round door out of it, then router out 1.5 inches all the way around the outer edge so i could fit in a steamed 2x10 board for a nice wooden inner and outer door frame ( jamb ? )  imset into the steel, the owner of the company thought it was a cool idea...  the white steel will also make the welded , black epoxied steel hinges and trim really stand out !

When i start on the round door, i will post updates here for ya.

While i am waiting for the delays in construction to work themselves out ( hurry up and wait ), i thought i would do some research into the door hinge on my 7 foot round door - that is 10 inches thick !  ( I love the R42 rating  )

Now, i will say, i am handy with a grinder, welder and epoxy, so if someone out there can help me with the design of said hinge, there is a good chance i can build it. - would like to include ball bearings in there as well.

One idea, is too have 2 bars extend off to the side, with the hinge in line with the far side of the door - the longer the bars, the more "swing" the door will have...

There is also the possibility of tapering the door outside - in, to make room for said "swing" and by adding a frame trim piece on the inside face, it can be reasonably air tight for the winter...

Wonder how thick a door this hinge will open ?

Edit : found the other round door thread, will read up on that one as well.

Edit.  Added round door imfo to my hobbit home thread.
https://permies.com/t/180629/Hobbit-Home-Progress#1424614

Thought i would post this here as well...

Large round door was looking a little worse for wear, with all the flaking epoxy covering it, so it is time for some T.L.C, while i continue to wait for the concrete to be poured...

Used a scraper to remove the flaky bits, then belt sanded the whole surface with a 50 grit belt sander to roughen up the surface, dusted it all off, then gave it 2 coats of outdoor, oil based, flat black paint.

Since it was starting to drag a bit at the bottom, removed some previously installed small shims, and put in some 1/4 inch shims too tilt the door up off the threshold.  Which in turn moved the door closer to the opening side of the door.  Painted the inner jamb of the door as well, so i could see the 2 spots where it was dragging - brought out the belt sander again, and sanded of a few millimetres in those 2 areas, then repainted.

It is back too being a nicely balanced, nicely closing door... that looks good as new.

Well, I learned a whole set of new concepts here.  This thread was opened for additional thoughts, so here are a few lame brained ones to chew on.  A taper is a good idea as it helps seat the door and actually will make up for very small changes in connection.  Just a thought - why set the door in a jam in place?  Perhaps this door jamb is already built, so this will not count, but - if the door has not been put in place, it would seem easier to build the entire door and frame in one horizonal plane on some heavy supports.  I would suggest this could be immediately in front of where you wish to place the door with a temporary roof to keep weather off.  Anchor the supports well as they will be important for final placement.  Set the entire square piece up and cut the door with a slight angle wider to the bottom.  Place the hardware for the door knob mechanism on at the half-thick phase of construction.  Using an offset double lock such as found in a bank vault.  Bolts can be allowed to slide both ways into the jamb with metal reinforcing around the edge of the door and the receiver.

Once the door assembly is complete, heavy plank or timbers can be connected on the two upright sides and the top of the door jamb.  Temporary holding wood pieces can be screwed across the door to keep it integral with the jamb.

Once all is complete, slide the entire assembly to the bottom front location for a door assembly.

Here is where the anchored supports come into play.  Lift the top of the entire assembly while sliding the base to its final resting place.  If pulleys can be set inside or overhead, then ropes can be attached that allow pulling of the door from the inside or just overhead.  Lift the door at the top with jacks and possibly timbers or small tree trunks with someone setting anchors at the lower end of the lifting timbers (keep anchors even and it is suggested there be at least two for each lifting piece.  This way, incremental pushes can be made with stops for rest and review of progress.  If the lower timber can be kept above the threshold until the door is coming to rest, it will be easier to lever the unit.  The entire door assembly can be lifted into place and secured with minor adjustments from a sledge.  This way, the door will remain in perfect alignment with the jamb and will be much less "ornery" to adjust.

Just a thought.

While ruminating on this project, I wanted to provide a couple of thoughts in addition to the above.  The first is philosophical.  There has been discussion on open in or open out.  I can tell that the works of Tolkien have influenced this project significantly.  The hobbits, which appear to the the driving force behind the project, were a friendly and welcoming people.  In that sense, it is logical to see their doors as being welcoming as well and opening in.  Now, from a physical standpoint, this, in a 7" diameter door means that a minimum sweep area of about 3,217 sq ft will need to be reserved inside the structure just to allow the door to open fully to a right angle with the door jamb.  That is a lot of room in most residential structures, the size of a decent room.  If the door opens out, that is a more defensive psychological message and was seen at the doors to KaZad-Dum where the dwarves were a less trusting race.  It takes much more to force doors that open out rather than open in.  Even if the door opens out, it will be potentially subject to more weather impacts as wind will more readily turn it into a sail.  I worked several years in the Rockies and expect that the Permies site in Montana is more flat than mountainous.  If so, we used to have a saying that the wind never stops in Wyoming where we had to operate most.  I would suggest any out-opening door be placed relative to any prevailing winds so they would push closed rather than open.  In addition, snow piled at the entrance by drifts may make opening that door more difficult without mechanisms to assist the process.

I see that a design seems to be trending toward a 2" thickness rather than a 3" one.  This would tend to lead to the conclusion that a 7 ft diameter round door of white oak would weigh, without hardware, around 302 lbs.  A 3 inch thickness would run about 452 lbs. That is massive indeed.  When that vertical weight is added to the horizontal moment created by the distance from the center of gravity to the hinges and the fact that most hinges end up required to handle the entire weight of both vertical weight and moment force, it is suggested that heavy duty commercial hinges capable of handling around 600 lbs be considered.  

A true round door will have a very limited threshold one can move through without tripping regularly.  However, if a wider threshold is called for, it will be necessary to make the door larger than 7 ft in diameter to allow the taller folks such as Paul entry without chancing a sore head.  He would likely not like my old farmhouse as it was built with energy efficiency rather than grandeur in mind.  Our kitchen ceiling is less than 6 ft 5 inches and the upstairs bedrooms are closer to 6 ft.  Luckily my grandmother lowered the family height several inches without realizing how useful it would be.  My great-grandfather was said to be 6 ft 7 inches in his stocking feet while my father and I came in at just over 5 ft 8 inches.  More hobbit than Numenorean.

I would like to throw out a thought, to reduce the sweep area and the potential for weather impacts (I saw something indicating a 5 ft wide overhang that would only cover slightly over half the 7 ft diameter door when open out) and wind is a given.  That would be to consider if a heavy rail that allowed the door to open inward toward one side or the other might be considered.  A rail could be designed and installed that would allow the door to be pushed in on one side and then roll inward and along the inside wall.  This remains a welcoming door without sacrificing quite so much interior space. It also can be controlled along the bottom by placing side rollers obtained from roller skates or skate boards to hold the bottom steady during operation.

I hope that a video is made of the final design and successful operation.  I love the concept, it is just my Scotch nature that prevents my attempting a duplicate design.

Richard Henry wrote: There has been discussion on open in or open out.  

I have learned - by actually living with a 7 foot round door, that yes indeed, the wind does catch it with some force, so yes, i am still debating on the door opening in - or out ?  Maybe install one of those small electric radiant  heating element pads in the conrete too melt snow so you can get out ?  Only turn it on when needed ?

Also, since i will be using the foam panel that is cut from the wall panel for the door itself ( 10 " thick ) i also have the option of off-setting the entire door into the house 2 inches too make room for hinge connection points inside as well as waterproofing and drainage options outside.  This will lower some insulation value at the edges of the door, but it is still far and above the R value of a standard door.

I am also keeping in mind, it will be located in a sunken front porch - 5 feet down with an earth covered roof overhead.  Should reduce the wind force a bit ?

LOTS of options... lots too consider.

Dave:

All good thoughts.  To control the force of wind on the door either way it opens, you may want to consider an old country trick to keep a dog from snapping its chain.  Hunting dogs, when young are not happy to be on a chain.  Some learn how to run just right that they over stress the chain and snap it if they get a good pop on it.  We used to take an old bicycle inner tube and tie it so that the tube would have to stretch pretty far before the chain came taut.  The spring in the rubber is just what is needed to slow the final snap on the chain.  That works for anchors in wood as well.

To fill old screw holes, it is easy to take some good wood glue and smear it over a wood golf tee and drive the tee into the hole.  When the glue ripens, the wood will be about as solid as it was when first installed.  New screw holes will hold fine.

I do not remember heaters being all that efficient on snow and ice.  I would consider a more environmentally safe product such as "Bare Ground" which is a liquid.  It uses corn as a base and can even be diluted in half to protect sensitive vegetation during an ice storm.  It is heavier than water, so drills through ice in temperatures up to -35 or so.  https://www.bareground.com/product/winter/mag-plus-liquid-deicer/ I have used it on a sidewalk the neighbor kids drove over with pickups and pounded the ice to near metallic density.  It would take about 15 minutes to cut the ice right off the concrete in near zero F temperatures.  It is pet safe and non-corrosive.  I actually used it to cut ice off my windshield although it would get a bit sticky later.

You are going to have a sunken front porch in Ontario, Canada?  I hope your overhang is sufficient to keep snow well back from the door.  Holes often tend to fill with snow when the wind is from the right direction.  I live in upstate NY and remember drifts along the low side of a hill reaching more than twice my height.  A hole tends to create a natural drop in wind speed.  Since the energy of wind as in water is the mass times the square of the velocity, it does not take much of a velocity drop to allow whatever is in that wind to drop out, be it snow or dirt.  The wind energy to carry drops off like a rock when the wind slows for any reason and a hole is a reason.  At the very least, the snow will develop a cornice on the edge of the hole which may allow snow to cave inside if it overloads the overhang.

You should understand that a sunken porch will allow colder air to drop and hang around there.  When building a snow shelter, it is suggested to dig into a drift and create the sleeping area above the entrance to allow warmer air to stay inside and colder air to drop down and flow out.  Another consideration is if you are using a combustion process inside the structure.  CO2 is heavier than oxygen and a sunken building using any heater that consumes fuel and gives off CO2 may cause elevated levels, especially during the winter when ventilation is lower.

Another concern is

Richard Henry wrote:
You are going to have a sunken front porch in Ontario, Canada?

Yup.  Have some ideas about avoiding snow drifting into the hole.
Not too worried about the cold settling into the hole though, since the front wall AND the front door are rated at R 40.

Oh, ya... this is gonna be an interesting build !

Not sure if this is interesting at all, and whether or not it's dumb as sh*t, but I had a random idea for an alternative "hinge" system that might work for a round door. What if, instead of proper hinges, you had a number of rails extending outward from the wall (or inward, if you want an inward-opening door) and the opening/closing of the door consisted of pushing the entire thing out or in? I.e. the door would slide on the rails, but always remain parallel to the wall. Of course, you'd have to keep quite a large area free from stuff in order to keep the door openable, and you'd need to keep the rail system well oiled or it would make a hellish noise. Also, it might look more like something from the Myst game series than from The Hobbit. I don't know nearly enough about this kind of stuff to say for sure whether this would be doable, or be better than hinges in any way if it was, but it seems to me that you might avoid the issue of the weight of the door warping the hinges...

Please ignore this if it's stupid.

No idea is stupid.  You do have potential, but a couple of things would be required to make it work efficiently.  If the door slid outward, the bulk of the door would reduce wind entrance to the inside, but also access for anything you wanted to get into the structure.  The rails would have to be long enough to sufficiently allow maneuvering around the door and into the structure.  This would require some horizontal space on one or both sides both inside and out to allow the diagonal entry.  I believe the horizontal space requirements would hold for open out or open in unless one were able to use pivoting trucks to roll along the rails with curves to move the door to one side or the other.  The lower rails would have to be inset into the floor to prevent a trip hazard and the upper rails would need to be quite tall to prevent folks like Paul from whacking their heads.  With insetting, the lower rails would be stable, but keeping the upper rails aligned and stable might be a challenge requiring robust connection to any inside/outside roof system.  The weakness of inset rails would be the need to keep the rails from holding debris that might block the trucks.  Any pets might have to be careful not to drop into the space and sprain a foot or leg as well.

The proposed hobbit home for this project would have the additional challenge for an outward door with respect to resistance to snow piles at the entrance.

Where your suggestion really gets my mind going is with its "out of the box" thinking.  What about placing the hinge on top of the door and using a cable or hydraulic system to hoist the door?  That seems to reduce the issues of uneven forces on the hinges although a number of other thoughts follow that.  Expect no one who is really into hobbit homes is looking to have anything other than a swinging door, although the permutations of entry are numerous.

My out of the box would be to build a square door in a round frame.  The door could be set on trucks in the entry with a sliding threshold that would hide the rails when open.  The bottom corners could be made to look round but open to square which would significantly reduce trip hazards, especially when partially open.  This would seem more efficient in that the door would slide into the frame on one side or the other which would provide ability to open partially or fully with no area needed on ether inside or outside for the door to sweep, would be naturally resistant to wind and provide a much more positive seal with no issues from out of alignment hinges.  All this while projecting a pure round door to the world and even to the inside should one wish.  The handle could remain in the center with a recessed groove in the frame where it could slide.  Latch could be up or down (or both), but security would be unmatched.

This has probably been mentioned above, but if you don't need a truly round door but you want it to look round, just make a round door and cut it vertically at the 1/4th mark from one side.  Have hinges on that vertical line so the remaining 3/4s of the door swings easily/normally.  Bottom threshold would still be round.  If the joint was disguised well enough, most people wouldn't know it's not truly round.  And the swing would be less obstructive.

I thought about hinges at the top.  Lots of benefits but if the mechanism fails, you could squish someone...

Mike Haasl wrote: most people wouldn't know it's not truly round.

Exactly the effect i was going for when i cut out the shed door...  from a distance, it looks truly round, but has a 36 inch threshold with a 36 inch hinge.

I am feeling a bit foolish as I just realized that a really good way to handle the swing of a large, potentially massive round door would be to use the same mechanism as a backhoe.  Hydraulic cylinders mount to the main boom (door) and swing on a top and bottom cast steel mandrel set in grease-fitted bearings.  There should be tons of old backhoes that have seen better days to salvage the parts from.  Backhoe booms often swing 180 degrees and can handle wind or any other forces.  I operated backhoes for over 20 years when I was starting out and should have thought of that as an option right off the bat.  A hydraulic pump can be driven by an electric motor to provide the force to run the system.  It is also possible to make a hidden handle that could work the system manually.  All one needs is to have a pump that could run the hydraulic fluid.

I'm really enjoying reading about the round door and the brilliant approaches to deal with the challenges that are far beyond my skill set.
Not to go too far off the topic, here's an enchanting diversion for those of us who lack the strength and equipment to put all that weight on side hinges. The central pivot approach might keep the round door visions alive for an interior application. While the approach doesn't conform to the official rules, and wouldn't work on the Wheaton Labs building, it could work for someone reading this thread who wants a round door with a lighter profile. For a less robust, non-sealed application, consider modifying the hinge location to the central pivot design shown here as in this arch door that could be reimagined as a full circle:

Source

This is an elegant solution so long as one has a full width threshold.  On a round door, however, the only portion of the threshold that would be low enough to provide a safe, trip-free passage would be taken up by the center post/pole/rotation/center.

One potential way to get there with this concept would be to place an eccentric above the upper jamb and possibly (for those who like balanced system, below the threshold) that would move the sides so that the round sections on each corner move with the door.  This means that the right hand circular corners would rotate automatically inward if the door is entered on the right and the left hand circular corners would rotate automatically inward if the door is entered on the left.  Of course, this means the doorway no longer is purely round, but it would be both safer to operate and allow easy entry while carrying something.

Roughly the same motion might be achieved by using bicycle sprockets and chain.  I anticipate that some rounding along the join line with the door jamb might be needed to allow movement each way.  The door could not easily pass a 90 degree angle.

The round door at Allerton Abbey mgith finally have the right hinges!  Learn more: https://permies.com/wiki/215113/Podcast-Office-Cleanup-Updates-Part

First attempt at 10 inch thick, R42, double weather stripped, mini round door test build - failed.
Used 1x6 and 1x4 for steps - will try 2x6 and 2x4 too increase door clearance.
More details here :
https://permies.com/t/240/180629/Hobbit-Home-Progress

Dave Lotte wrote:First attempt at 10 inch thick, R42, double weather stripped, mini round door test build - failed.
Used 1x6 and 1x4 for steps - will try 2x6 and 2x4 too increase door clearance.
More details here :
https://permies.com/t/240/180629/Hobbit-Home-Progress

Fascinating, Dave.  So you're going to adjust your jam clearance?  It looks like a similar design to what we're going for here, in regards to the multiple nested contact points.  

I think we're going for a wooden door at the lab.  Did you use SIPs?

Beau M. Davidson wrote:

Dave Lotte wrote:More details here :
https://permies.com/t/240/180629/Hobbit-Home-Progress

So you're going to adjust your jam clearance?
Did you use SIPs?

10 inch thick R42 structural insualted panels .... yes ( s.i.p.s )
Tinkering with it.  Did not feel comfortable just chopping into my brand new front wall without doing a few tests first.
Increasing the size of the steps just to get it to open and close.
After that - open and close - AND - match the door frame.
One issue i have to keep reminding myself about .... this test door is only 18 inches or so wide - it will be a bit different on a 84 inch door.
Feeling more comfortable about cutting into the front wall now though.

Hello Everyone!

Its been a while since I've done any serious wood working, hopefully the last ten years spent welding and machining will carry over to working with wood.

After a little deliberations, I think the largest challenge to this whole project boils down to the shear size and weight of the door, being necessary to have a minimum 6' diameter to comfortably step through, which causes deflection of the hinges as the load levers down. This necessitates not only extremely rigid hinges, but an equally stout frame to support this moving, swinging mass. The intertial moments acting upon the hinges are very high. This creates a another consideration, the accuracy of the mating flanges. Because of the inherently higher deflection owed to the greater elastic properties of natural materials such as wood, setting the hinges in such a way that the door seals properly will prove to be difficult. Not to mention swelling and contraction with seasonal cycles of humidity and temperature! An adjustment mechanism would almost certainly be needed, the probability that anyone could get this dead-nut right the first time is very low.

There is a way to balance the loads and forces, however. If we hinge the door at the top, and use a steel cable or chain to open and close the door, this allows the forces to bi distributed equally and assures a good and accurate fit for a good seal/closure of the door. This eliminates the need to engineer a heavy duty and adjustible steel hinge pins and system that can adjust for lash and deflection, and allows for a more symmetric and easier design and construction of the door frame. Speaking of the frame, it will prove to be a greater challenge than the door itself. The door is basically a wooden disc, all the engineering and design is in the frame and hinge system.

As crazy as it sounds, using wood in this way presents even more challenges than an engineered material such as steel. A hydraulic blast door or bank vault door would be just as approachable to me as this project, but I will do what I can to see to it that my promise to Paul is fulfilled.

So if it pleases the crowd, what are everyones thoughts on hinging the door at the TOP rather than the side? The only downside I can think of is that if the hinge or lifting chains or their attachment points fail, the door could fall and cause injury, however the same could be side of a conventionally side-hinged door (albeit I don't suspect the fall would be as violent). This could be mitigated using a torsion or tension spring so that the weight of the door itself will large be lifted by the spring, creating a balance between the spring force and gravity, with human strength being used only to open or shut the door. This is how rolling garage doors function, they have a torsion spring that does 95% of the lifting.

So I have this rough model of the frame. The weight of the frame alone is going to be over 600 pounds. This will easily be the most difficult carpentry project I've ever taken on. I'm really liking the idea of hinging it at the top and using chain to open and close it.

We're going to need to figure out how to attach it to the foundation. Concrete would be ideal here due to the weight involved. Those vertical beams are 8.5" X 6"

Sky Huddleston wrote:Those vertical beams are 8.5" X 6"

Good too see you making progress.  Second attempt here is in the works,  just have to do a bit of cutting tonite.
Not sure how much my finished door will weight in at, mostly steel laminated foam with (1) 2x10 - (1) 2x8 and (1) 2x4 around it - plus the hinges.
The big thing is, getting the experience working with the foam materials before the main event.

Updates and further deliberations.

After more consideration of construction and design, I've concluded that the hardest component to this is actually the mating surface the door fits into. The structural support frame, if it could be made square, would be very easy. So too would the actual round door itself. But if the door is to mate against a corresponding round frame, then that makes it 10 times harder easily. This makes me wonder if we could use a square timber frame for the structural support, a round door, and use cob or adobe or similar to actually create the mating area/seam for the door. Without doing something akin to this, the project may be too difficult to take on using the specific natural materials Paul wants.

What are everyone elses thoughts on that?

If you are hinging it vertically, you could make it open down instead of opening up. Like a drawbridge. This still leaves safety considerations. But the people passing through do not have to go under a very heavy door.

Sky Huddleston wrote:I've concluded that the hardest component to this is actually the mating surface the door fits into. The structural support frame,

Agreed.
This is why, i built my shed round door out of all one piece - full front coverage - sheets of plywood.  I believe i posted pictures on this thread....
By cutting the door out of the frame itself, you take the guess work out of the fit and finish aspect of the build, just add hinges.
With my foam door, i kinda have a guide - the steel is still providing the round shape - but there is a ton of wiggle room when applying the 3 layers of kerfed planking - for something to go wrong.
Of course - there is a ton of wiggle room - when applying the kerfed boards - for fixing it as well...
2nd test build drying now, but it is difficult working on something so small.

Sky Huddleston wrote: Without doing something akin to this, the project may be too difficult to take on using the specific natural materials Paul wants.

I have done a variety of projects in my time,  and have a couple of friends who look at me and ask
" how do you do that ??? "
May i suggest the same thing that i tell them ?
Start small, and with each victory - it builds up your confidence and knowledge base for the next one.
Don't try to build a 7 foot round door all at once - build a 2 foot round test first.  Same materials, same thickness, just .... smaller.   Call it a doggy door.

I have also met people who never get past the planning stages as well,  the project just .... stalls ...
go out and do the mini.  😁

Dave Lotte wrote:

Sky Huddleston wrote: Without doing something akin to this, the project may be too difficult to take on using the specific natural materials Paul wants.

I have done a variety of projects in my time,  and have a couple of friends who look at me and ask
" how do you do that ??? "
May i suggest the same thing that i tell them ?
Start small, and with each victory - it builds up your confidence and knowledge base for the next one.
Don't try to build a 7 foot round door all at once - build a 2 foot round test first.  Same materials, same thickness, just .... smaller.   Call it a doggy door.

I have also met people who never get past the planning stages as well,  the project just .... stalls ...
go out and do the mini.  😁

Right now I don't have time to commit to that. I have multiple development contracts for rocket stove based gas turbine engines and another contract for the development of Bourke engine technology. The round door project is basically going to have to take not more than 3 weeks, tops.

If there are any carpenters on the forum that I can pay to build the door according to design blueprints please let me know.

Talked to Paul.

No top hinge.
No concrete at the Lab.
Make it look like a classical hobbit door - wooden door, knob in the middle.


My thoughts:
You could use timber frame, but you could also engineer metal components to frame, metal door bracing, and latching mechanism, down at the shop, then take it up to the lab to partially demo the wall, install suitably beefy timbers for strength, and install your metal structure.
You could then manufacture the door around the metal skeleton using cross-laminated lumber and wooden dowels.  

I don't think cob where the door seats against the structure will be a lasting material choice.  

does all the round have to open?  I really like the way they set up this door in the hobbit village.



from https://www.hobbitontours.com/experiences/hobbiton-movie-set-tour/  

Admittedly, I'm a lazy builder and find hanging doors to one of my biggest bugbears.

Although, if it needed to be round, I would go with a hinge these ones as it would hold the weight nicely.



It wouldn't need to be as pretty.  

Per Paul, yes, the actual door must be fully round.

I think beefy metal bracing and hinge is a good possible path.  Which happens to be Sky's specialty!

Great hinge!  Sturdy and gorgeous!

Nice looking hinge, but where would i put my sting swords and dragons head door knocker 😁😁😁