Bumping this because it looks like the best analysis of dimensions we have on a police box.  I mean, it may not be the most detailed, and it may not delve into Mk II, III, IV variances, but it's a simple, close enough for government work analysis.  I'm thinking about trying to give my Barachaki-based TARDIS a 3 stepped roof with the shorter capitals on the corner posts and the measurements here help me get a feel for if I can make it work...ish. 

These are the drawings that I used to build my TARDIS/Police Box. I couldn't have done it without them. 

Greg.

Yes, these were a rather marvelous first attempt at digesting the Trench and GPO plans and transforming them into useful measurements, filling in the missing details using real-life measurements from Crich. It's a shame Anita's efforts didn't progress to a full set of plans.

As we've discovered in the intervening years, however, the results from this method of reconstructing things can't help but be a bit of a gestalt of all the concrete Met Box models (excepting the oddball Mark 5's of course). There's nothing wrong with this, in fact my/our own TARDIS was distinctly a gestalt affair (oh, how little we knew back in 1986 ), and in many ways it can be a great way of evoking the multiple variations over the years.

Nonetheless, the limitations do have to be recognized. First, you have to consider where and when the undated Trench plans came from. With what we've learned over the years, on morphological grounds alone it can now be seen that they are, on their face, working drawings for the purpose of soliciting tenders for the first Mark 3 Met Box installations in S and X Divisions, which were actually erected in mid-to-late 1935. As such, they probably date from late 1934 to early 1935.

The reasons for these deductions are many fold. First, of course, they are for a concrete Box rather than a mostly timber one like the Mark 1's. We now know from various memos and other internal documents that the evolution of the design was progressive and organic over time, changing with the needs of the project. When it was in its experimental stages, and even a little after that, the emphasis and demands from the higher ups were for doing it on the cheap and with as much flexibility as possible, both in design and mobility of the final units. Timber fit the bill perfectly, so it wasn't designed out of the gate as a concrete design only to be "dumbed down" for the timber Boxes (now known to be far more plentiful than first thought - besides the 10 Prototypes, populating 3 full sub-divisions with a smattering of a few more Mark 1's elsewhere). Instead, the use of concrete came only after it was decided the project would continue with a full head of steam, when the need for durability and the ability to spread the greater setup costs over more units made it more economical.

The design of the Boxes also show a distinct, and internally logical progression from Prototypes, to Mark 1s, to the first all-concrete Mark 2s, the improved Mark 3s, the war and post-war replacement Mark 4s, all the way through to the simplified Mark 5's beyond, with the differences from one to the next showing an arc of evolution.

The Trench drawings also show signs of having been redrawn from an earlier set. They are sometimes internally inconsistent, and many of the shapes aren't precise (eg, the roof plan isn't even square), and if you check the dimension call outs against the scale they are often at odds with each other. If it weren't for those call outs, in many cases we would get very different dimensions for a Met Box. The roof stacks even change a bit from view to view, some looking more like Mark 2's, sometimes more like a Mark 3's, and one even with slight reverse stacking like a Mark 4, despite the call out saying they were supposed to be exactly even.

Morphologically (ignoring the drawing errors), some of the views show distinctly (early-to-mid) Mark 2 features, particularly in the elevation views of the top sign lintels (which look like they are trying to represent surface mounted sign frames instead of inset ones) and in the roof stacks, as well as the way the entrance door is constructed. Those sign lintels also show indications of an idealized length for them, ending in the middle of the sign posts like the Mark 1's did, and then the Mark 4's later on. This leads one to believe they are a redrawn set from the original Mark 2 plans (which also formed the basis for the Glasgow Boxes), amended to incorporate the later Mark 2 and proposed Mark 3 improvements. Other parts show obvious elements of Mark 3s, including the window pane patterns and which ones open, the separate detailed cross-section for the sign lintels and the lack of vents in same, and most definitively, the outline of the telephone used in the plan cross section is without doubt that of a GPO Telephone PA No. 7, which didn't even exist before the PA 150MP signal system was developed for the Met in late 1934. Then there's the interim wording of the top sign as "POLICE * PUBLIC", which seems to be a work in progressing, being more than a Mark 1/2's simple "POLICE" and less than a Mark 3's full "POLICE PUBLIC CALL BOX". There are even more signs of later experiments with signage on this sheet, with what looks to be a much later addition of roughly hand drawn wording on the Phone Door despite the older plate sign being specified and drawn beneath (said plates originally being retrofitted on the first Met Boxes in late 1930 - not even recognized as being needed before then - and remaining in use throughout the System until late 1937).

The other document used, the GPO drawings that were drawn in 1937 for the PA350 system then in development, brings it's own idiosyncrasies into the mix. They seem to be essentially of a Mark 3, which was the current Box at the time, but with its own set of idealizations, and others that seem to hearken back to the Trench plans.

So, all of this means that taking the plans as a start will, off-the-bat, tend to produce a gestalt of Mark 2's and Mark 3's, with a few idealized call backs to Mark 1's and some foreshadowing of Mark 4's. Add to this the measurements from a (much repaired) Mark 4 to fill in the gaps and the stretch across the design spectrum is lengthened even further.

It is, without doubt, great for what it is, but it's really just a starting point. Ideally we would have a full set of reconstructed plans for all 5 Marks of Met Box and their major variations within those Marks as built, along with as many of the Prototypes we can find and the earlier Victorian/Edwardian model which preceded them, and even the homage that is the Earl's Court Box. To this would have to be added a full, corrected redrawing of the Trench plans as an idealized, interim version of a Mark 2/3.

So, where do we go from here?

Well, I'd certainly like to help in whatever way I can. I'm definitely no architect, nor a builder, but if needs be I could certainly make some plans from my models to work as a basic starting point, even if it's just to get the overall shape right. Generally I make mine from overlaying the models onto head-on photographs (and if they're not squarely head-on, I adjust the perspective in Photoshop to make them so - this can sometimes cause its own host of perspective issues in itself if not done correctly, but is better than not doing it at all).

As an example, here's my Mark 2. It's probably not very useful without actual accurate measurements (especially since we don't have any measurements of a Mark 2), but as a proof-of-concept thing it should work alright.


Arbitrary measurements could potentially be worked out using actual photos, I suppose?

Good first stab, lymerence, but perhaps not quite ready for prime time. We might want to get things squared away a bit more first with comparisons to photos, etc. in the 3D model stage, both visually and via photogrammetry, before committing to the plans/measurements stage. Just a thought - though we could do that here just as well.

I was going to post this to your latest Mark 2 build last night (http://tardisbuilders.com/index.php?topic=8412.msg109414#msg109414), but got distracted.

There are a number of small issues that I'll get into more detail about when I have the time tonight (probably there, with a repeat or link here), but the biggest is that the early Mark 2's with the globe beacons did not have any kind of square base above the roof. I'm prepared to prove this assertion to a reasonable degree of certainty, as I went back last night to double check my earlier observations. Note also that there should be a small cylinder or pipe coming up through the peak of the roof to the underside of the globe, which was likely both lightbulb fitting support and electrical access to same.

Two other quickies - your top sign attachment frames are too thick, as in too much is showing in this view (and in your renders they also had the wrong texture, as they were almost certainly metal/steel) - and second, your lock is too far to the left, should have about 1/4 of the lock to the right of the rail/stile join line. You should probably show those lines for the front door as well, along with the hinges (which on Mark 2's were only 5" as opposed to Barnet's 6").

Other observations are that the pillars aren't quite high enough (overall top should come about midway up the first step), and the amount of inset panel surrounding the phone door cut isn't quite enough (meaning the phone door is a bit too large as well). Oh, and the metal parts of the windows aren't quite thick enough (in this view - not talking about thickness proud of the window panes).

Note also that while you've got the roof tier sloping flats about right, you haven't added the slopes to the sign lintels (though even Trench didn't show that). They sloped quite a bit up top to shed water (on par with the roof tiers), but I believe below as well just a bit for mould release reasons - then they flattened out for the rest of the ring beam area you can't see. They were two-piece construction, with the main part of the lintel up to the level of the flats of the pillars being part of the continuous ring beam that held everything together up top, with the fascias that stood proud of the pillars being bolted on top of that/them.

One last that escapes most people's attention - these early Mark 2's had vents on the undersides of the side sign lintels (two each lintel, but only the left and right sides at first, then later often back and even fronts had them as well due to a systematic build error that only came to light when they caused issues in use). Again, I'll show photos later. Note that Mark 3's and 4's had no such vents, as more of their windows opened and there was some additional air flow out the roof through the lamp housing.

Again, more later.

I am rather excited though, that you've added the energy, time, skills and additional set of eyes to start moving in this direction - so kudos!

No problem! I'm honestly always glad to have feedback like this - hopefully at some point they'll be accurate enough to be useful. The plans posted earlier were of course just a proof-of-concept (mostly as a test of rendering the model as a plan). When I'm dedicated to something like this, I never know when to stop.

But I'm glad to be of help; someday I hope I'll be as knowledgeable as you on the topic. We know a lot more about Met boxes now than we did even six years ago, but there's still always a little bit of mystery surrounding it all, and I'd love to be a part of that research (I've just ordered The Rise and Fall of the Police Box too, so that should make for a good read). I never expected myself to get this invested in it, to the point where I'm excited by even just a new photograph, but it's a great community and I'm glad to be part of it.

They've already been very useful, with the promise of being even more so.

I had long ago intended to do something like this, but time never seemed on my side, so your contribution has been a much needed boost to things.

Bunker's other book, "From Rattle to Radio" is also a good read. Only most of a single chapter on the Met Police Boxes and their antecedents, but still interesting, especially if you have any interest in the 1897 boxes, as there's a bit more info there on them than in "The Rise and Fall..."

BTW - The Trench plans may be a very good place to start with measurements on the Mark 2's. As noted above in my last post before your first, they seem to have been traced from one version or another of the Mark 2 working drawings, then just modified to reflect the planned changes for the Mark 3's (as with the Glasgow Mark 4 Boxes). As such, some of their measurements, especially in the roof area, seem more applicable to the earlier Met Boxes.

Of course, not all the measurements are called out, nor are they drawn with enough accuracy to be able to reliably read off measurements from the given scale (drawn details often differing even from one side to another on the same elevation or plan), but if one interprets things to more or less the nearest even fraction of an inch (1/4" or 1/8"), I think it's a good place to start.  Then line them up in wire frame to good photos of various Mark 2s and see where things fit and where they don't - adjusting as necessary, just like Tony Farrell and dw_1200 have been doing so brilliantly with the Brachacki Box (http://tardisbuilders.com/index.php?topic=8649.0).

If I could just 'chip in' on Alan's final point about overlaying wire frame diagrams on photos, be prepared to have a great many goes to get it right.

Slava (DW_1200) and I are still working on this but rather than boring the socks off everyone showing each minute adjustment, we decided to do the work just between ourselves with numerous e-mails flying backwards and forwards between the Duchy & County Palatine of Lancashire and the imperial city of St Petersburg!

You'll be surprised too how what might be thought of as a minor adjustment to one single dimension - even less than an eighth of an inch - can throw everything else out. As Mike Verte has demonstrated in his rather nice little videos, straight lines and perfect angles don't really exist in the real world. This fact is particularly true of the Brachacki box - as early as Season two through constant man-handling (and a couple of serious wallops) the Tardis has started to twist out of shape and by the time of "The Gunfighters" the old girl is on the point of collapsing altogether!

The same thing is going to be true of concrete. Even if the parts were all drawn from the same moulds (which, if I understand correctly, they weren't), constant use of those moulds would result in the moulds becoming distorted and having to be repaired. Then you'd have to ask are the parts all put together with the same degree of accuracy? Is the box you're studying actually level on the pavement? Then there is maintenance: If the door has been repaired, has the door been re-hung at the same height or have the workmen sanded a bit off to get it to fit?

Have fun with this project Lym but, be prepared for some hard work and head-scratching along the way!  

T

Thanks for all that Tony, and I was wondering why things had gone quiet after so much interesting activity.

It's actually even worse for Met Boxes after the Mark 1's, though. Not only were they designed to shed water, so with almost no right angles in a vertical sense, they were also moulded in concrete, which means that many parts couldn't be at right angles due to mould release issues (to say nothing of the wide tolerances inherent in both the materials and the moulds themselves).

I think that Matt was quite surprised when he started making measurements at Crich as to just how irregular that structure really is - as I recall, saying at one point that he didn't think there was a single right angle on the whole Box. Even many of the non-concrete parts were also moulded, with all the same issues (though a bit tighter tolerances).

And as to construction, the loose tolerances necessary to be able to slap them together with just some mortar to fill the voids is obvious in many photos. One look at the sign lintel fascias on most Mark 2s and 3s alone tells a story, as you can almost instantly see that they were rarely anywhere near centred, with often huge variance in the distance to the end of the pillar between left and right sides.

Quote from: hb88banzai on Apr 21, 2018, 12:11 am
Note also that while you've got the roof tier sloping flats about right, you haven't added the slopes to the sign lintels (though even Trench didn't show that). They sloped quite a bit up top to shed water (on par with the roof tiers), but I believe below as well just a bit for mould release reasons - then they flattened out for the rest of the ring beam area you can't see.

Speaking of slopes, I noticed the base of the box in both the above drawing and in the renders has a sharp "bevel" around the top edge (like the TARDISes have). However both the GPO and Trench plans show a much shallower "slope", like that on the Top Sign boxes... at least in the cross-section drawings; most of the overall external drawings show no slope, just a plain square base.

So whether it's a slight slope, or no slope, the sharp bevel on the drawing and renders needs some flattening out.

I hope this is a little helpful.

Dino.

Yeah, it's a little steep. This is a detail that varied a bit over time and type, but I don't think any were quite that steep.

Here's a pic of the Ashcombe Avenue early Mark 2 Box at full resolution, cropped to the base --



Note that the top right corner of the base has been knocked off along the bevel line and extending below it, making it look steeper and deeper than it was.

Here's a crop of the later photo of it from that side, showing the damage there and also more on the same side at the back lower corner, where a big chunk is missing --

As for slopes, as mentioned earlier, the Crich box had them everywhere for rain dissipation however the base looks flat upon first inspection, certainly no chamfer like the two last pictures or the TARDIS.

If you look closely at it though, there is indeed a very shallow slope but nothing like these.

Having made a replica, I can confirm the Base on the Crich box does have a slant, but that the slant goes all the way to the wall, with no horizontal part.  So, the columns actually land on slanting surfaces, instead of flat, which makes it rather harder to build!  (But presumably, is more weatherproof, as there's no flat area to accumulate rain water...

Yeah, I figured as much. I have a sneaking suspicion that the seemingly flat section on the Mark 2 and 3 Boxes were also at a slight slant from the walls.

Here's a crop from one of Mark's photos to show the Crich (Mark 4) Box base --



Note that the edge of the base is also closer to the pillar than on the earlier Met Boxes. I'm wondering what the overall width of the base measures.

The Mark 1's also had a continuous slant from the wall to the edge of the base, but at a significantly steeper angle --

This is all great stuff! I've got a document in which I'm taking these notes which I'll be able to cross off when I add them to my models and plans. I plan to read through every topic posted to the Police Box section of the site (including those on the Met box catalog) and try and learn any other things of note.

It really is such a niche interest, which is probably why it's taken so long to try and organise accurate plans. If there's anything else I could help with even a little bit, just let me know.