My Garden Room Build - 9m x 4m

UKworkshop.co.uk

Help Support UKworkshop.co.uk:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Apr 2021 - Beam Calculations (for above the doors)

I spent way too long looking into different options for door headers. I was a little paranoid about deflection as I didn't want to have any problems with the doors working properly over time. I have heard of a few people that have had issues with bifold doors and the root cause seems to come back to a combination of:
1. the door header not being suitable for the door span + roof loads
2. the sensitivity of the door to vertical deflection (but reason 1 seems to be the big one)

In addition to this, changes in the weather through the seasons can (I have heard) cause flex and movement which can lead to issues. My doors are alu so that means theoretically more stability versus PVC, and hopefully less issues, but the main reason I selected alu was for the sharper looks (if I'm honest). So... operating from the assumption that vertical deflection is the true nemesis of bifolds this became the factor which I needed to focus on and my mind was set on giving my lovely new doors a stress free life via a suitably deluxe door header experience. This meant doing a bit of research into what my best options were.

Results from my research
Unfortunately, most of the good guidance I found was for multi storey buildings and hence they were often coming up with solutions that would be not be suitable for a garden room. For my particular application, these recommendations would be either:
1. overengineered in terms of strength and cost
2. cause me issues on overall building height
3. or both

Mildly relevant distraction #1:
Regarding the height thing, when working within permitted development you need to keep everything under 2.5m (when building within 2m of the boundary) and for that reason you really need to think about what you are doing every step of the way. For example, assuming 2000mm for your door height, that leaves only half a metre for the base, the headers and the roof (not to mention the air gap under the building, and possibly even your floor boards depending on how you build it).

Anyway, I'm getting distracted, the point is the only other guidance I found for header selection / header design was pretty anecdotal and didn't seem to have any theory behind it. That isn't necessarily a problem, but on this occasion it wasn't giving me enough confidence to make a decision, considering the potential impact of getting it wrong (fiddly / sticky doors of annoyance).

The cunning plan
So... logically, I decided that I would leverage my complete lack of structural engineering knowledge to manually calculate the beam deflections for all of the options available. So here are the options I considered:

1. Doubled up 6 x 2, C24 timber
2. Flitch beam (C24 timber and steel 'flitch plate' sandwiched together)
3. I beam (sometimes known as a H beam or a 'universal beam')
4. RHS (rectangular hollow section)

Note: I did not include concrete lintel or glulam beam, for reasons which I honestly cannot remember, and there may be other options which I dont even know about, that I am hoping people will point out to me.

My plan was to use the below formula to calculate the maximum deflection on the beam. I would still of course have the problem of not knowing how much deflection was acceptable for my particular bifold door, hmmmm... HOWEVER I would be able to COMPARE them all

beam bend 1.jpg


Mildly relevant distraction #2:
I did contact some of the door manufactures to find out what sort of deflection they think their products could handle, or even what headers they recommended for different spans - but alas, nobody wanted to commit to anything specific, other than to say "computer says no" or "zero deflection required" which is of course theoretically impossible, even if I made it from diamond.
Note: making it from diamond would represent a significant cost save for the build, considering how much I'm planning to spend on materials and 'essential' tools so far)

Mildly relevant distraction #3:
Note that there are lots of different formulas for this sort of thing, but the one above should represent worst case for my door header because it assumes the beam is simply supported at either end (in reality it has lots of engineering screws holding it in place along some of its length and also has the roof load partially distributed either side of the 'fulcrum' too, which I imagine restricts the movement - not that I know what I am talking about). Oh yeah, please note that I am absolutely NOT qualified in this topic, and everything I say should be considered as science fiction, even the science part is pushing it a bit.
I will point out though that I am lucky enough to know somebody that does know what he is talking about, and he checked this over for me and helped me with some of the calcs, so it isn't completely worthless. As a minimum I think it gives a good indication of how the options stack up against one another. Worst case its a bit of fun, and a good opportunity for everybody to laugh at my mistakes.

Variables

beam bend 2.jpg


The legend in the pictures explains what the variables are, so I wont double explain those, but it is worth noting that the values for youngs modulus for each material can be found online in the so called 'blue book', as can the values for moment of inertia (which is specific to each shape / cross section) - you can calculate this if you want to, but its easier to look it up. I vaguely remember us covering how to do this in my mech eng degree, but that was 20 years ago and I was rubbish at it then and suspect even crapper now so much prefer standing on the shoulders of giants. Oh yeah, 'w' or roof load is the other funny one. I think for that one I calculated the weight of my entire roof (timber, insulation, EPDM, chipboard) and then added some for static snow load, and finally I think some for dynamic load for people / person (me) walking on there. I also divided the numbers by two because (approx.) half the load is on the front wall and the other half is on the back wall (my side walls do not support any load) Honestly can't remember exactly how I got that number, but as I say all my assumptions are clearly shown in the workings so as not to mislead anybody into how I got the overall deflection results. Looking at these numbers now, we have 14 N/cm which is 1.4kg per cm, which is 140kg per metre - so that 'feels' in the right ball park for the amount of load along that beam.
Finally, regarding variables, you will notice the slightly hilarious use of 'cm' as a unit. I am advised that the golden rule with SE calcs is to pick one unit and stick with it, and because a lot of the numbers provided for free in the blue book are in 'cm' it makes sense to use cm throughout.

How I will contextualise the results
Seeing as I don't know how much deflection is tolerable by the doors, due to the fact it is a secret (not disclosed by the door manufacturers), I need a way to contextualise any numbers that pop out of the calculations. I am clutching at straws a little here, but I did find some rough guidelines that I think could help put the deflection values into context: I understand that some people work the following rule:
• <2m span = doubled up 2 x 6 timber is ok
• >2m span = metal beam required (or you could use thicker timber, but you can't simply add in 9x2 headers when you only have 2.5m overall building height to play with - so we end up in metal beam territory once a 6 x 2 isn't strong enough)

So, I will use a doubled up 6 x 2 timber as a sort of baseline solution that is 'probably' acceptable for a 2 metre span. Considering my span is 2.4m, I know that any alternative designs only need to be say 20% stronger, for the same constrained / max 150mm header thickness. (I think I mentioned this already, but I cannot just simply add thicker timbers and be done with it, due to the 2.5m building height restriction. That in itself is the ONLY reason for doing all this faffing around - I am fixed with 150mm max for my header)

The results
I will include some extracts below from my summaries for each of the 4 options so that you can see my assumptions (mistakes) in all their glory. As you can see, the timber option (for my 2.4m wide bifold) gives 3.4mm deflection (in the middle), and the metal options are far stiffer ranging from 0.71mm to 0.34mm deflection.

beam bend C24.jpg


beam bend flitch beam.jpg


beam bend IB.jpg


beam bend RHS.jpg


Below are the full workings in excel.
excel beam calculations.jpg


The decision
So to cut a long story short (oops too late for that) I have decided to use doubled up 6 x 2 timber for above the french doors which have only a 1.6m span, and a flitch beam for above the bifold door which has a 2.4m span.

Factors influencing the decision:
COST
From a cost perspective timber is cheapest (about £50), and all the metal options work out in a similar same ball park, although flitch was cheaper than the other two metal options by about 30 or 40%
STRENGTH
Again, all the metal options are basically the same - as a minimum they are 4 times stiffer than timber, and if we work from the assumption that the timber option is good enough for a 2m span, then we can conclude from the numbers that the metal options are all way stronger than needed for my 2.4m. So, for my span, strength is not really a factor when choosing between the metal options.
WORKABILITY
the primary reason for choosing flitch beam over i beam or RHS was the fact that the flitch beam gave me a timber surface on both sides, which is nicer to work with (attaching cladding on outside or plasterboard on inside). I dare say flitch beam is slightly more hassle to erect than RHS or I beam because you have to drill holes in the timber and bolt it all together, which takes time, but its probably 2 hours of work to construct and install a flitch beam versus 1 hour to install RHS or I beam so probably not significant in the grand scheme of things.

Mildly relevant distraction #4: I just checked with future self, and I / he has confirmed that the flitch beam did take about 2 hours, not including the 30 minutes spent staring at it proudly afterwards, or the 15 minutes spent explaining it to my (completely uninterested) wife

Summary
Overall I believe flitch beam to be the best option for this application, but I reserve the right to change my mind later, probably when its too late. See below for some top secret pictures which I took from the future.

IMG_20210705_150903.jpg


IMG_20210707_152249.jpg


Cost wise we are looking at about £100 for the flitch plate, which then gets sandwiched between two timbers to create a flitch beam. Note that the delivery costs for these things are huge (prohibitively so), so I picked mine up in my A-team van as the place was only about 40 minutes away. You will also need some bolts too. And wood. Under £200 'all in' I guess, which feels like good value if it keeps the doors working properly.

I am sorry for the length of that post, I find it hard to filter and I also find it hard and distinguish between useful and useless information. But, one man's trash is another man's treasure. Maybe that means my garden room could actually be worth something in the future.... :-D

Martin
 

Attachments

  • IMG_20210705_160724.jpg
    IMG_20210705_160724.jpg
    176.2 KB · Views: 62
Looking good so far, those ground screws made me chuckle thinking about if they were phillips or pozi drive. Please do not stop posting detailed write ups of what you are doing,I find them interesting and with pictures they help to explain the process much more clearly(y)(y)(y):)
 
Brilliant, that’s exactly how my wind works and why it takes so long to get anything done. Overthinking from first principles as you have zero experiential learning to draw from.
 
Mark, I will do - thanks for the positive feedback.
Fitzroy, brilliantly summarised, that is EXACTLY what's happening! :)
 
May 2021 - Trench for Services

The goal with this stage was as follows:

1. Design a trench
2. Dig a trench
3. Install services in trench
4. Backfill trench
5. Dispose of excess dirt in the neighbours garden (in the style of The Great Escape)

So this was yet another topic which I know absolutely nothing about and hence needed to spend some time researching it.

Bathroom scope creep
Originally this was just going to be a trench for power and internet, but then my wife had the idea of having a bathroom in the garden room (doh). I admit that I resisted this option furiously to start with because I didn't fancy the added complication on top of all the other things swirling around my head with this project (pretty much every aspect was new to me at this stage). But after some thought I realised she was right, if I don't do this now, I will never learn, and anyway "how hard can it be".... (lol)
The other aspect to consider is that adding a bathroom adds a lot of value (so I'm told), not that we are selling the house right now, but for the 'small' additional effort and cost, it can make this project a much more convincing investment. But ultimately, it wasn't property value that was driving this decision, it was a combination of the following:
  • If we do have guests stay over on the odd occasion, having a toilet inside the building is a bit of a game changer in terms of usability
  • The right hand side of the building is being designed as a space for my wife's maths tutoring, and also as a social space to hang out in, so it's handy for those
  • Our house has only one bathroom at the moment so having another is a big plus

Trench design
So I have 4 things that need to go into the trench:
1. Water
2. Waste
3. Power
4. Internet (x2, one as backup or use for an alarm)

Below is the plan view showing the routing needed for each service. Internet runs from the front of the house (not shown), power and water (red and blue) from about the middle, and waste (yellow) is from a drain (inspection chamber) which is situated in the garden adjacent to the rear of the house.

Services 3.jpg


I had to calculate the amount of shielded power cable needed so it could be ordered, here is the sketch for that.

SWA Cable length calculation.jpg


Guidelines
Looking online, there is plenty of guidance (mostly in the building regulations) about the necessary depths of each, but not so much about how they all live happily together in a single trench. It seems that common sense is the tool of choice for figuring out the best placement of each item inside a combined services trench, in terms of how they sit in relation to one another. Here are the guidelines I was working from after a bit of research:

Depth guidelines (based mostly on building regs)
  • Electricty depth (from the regs) = 500mm (to prevent unplanned spade / cable interaction electrification events)
  • Water depth (from the regs) = 750mm (to stop it freezing)
  • Internet: set at a similar depth to the power to keep it safe from humans with spades

Proximity guidelines (based mostly on anecdotal evidence and / or common sense)
  • Data: keep data away from power to prevent any electrical interference
  • Water: Keep waste and water apart (just in case one contaminates the other, which would really suck admittedly, but it does seem unbelievably unlikely as essentially both pipes would need to be leaking for that to happen). But that is the advice I managed to extract from the internet...
  • Avoid placing power or data below water and waste if possible - if I do get a leak this seems like the best disaster to have from the two options, but in grand scheme of things I doubt if this matters in the slightest

Note: CAT6 cable is being used for data / internet, which is apparently electrically / EMC shielded, so its theoretically possible that power and internet could go in the same conduit. However, this seems to vary depending on who you talk to, some people are saying you need to separate them, and some people are saying you dont need to, even for unshielded CAT5. So I played it safe and put data in a separate conduit of its own, as far away from power as I could, and also used CAT6.
Note: by 'conduit' I mean plastic water pipe
Note: by 'people' I mean mostly people on forums, but also some tradesmen that I know and spoke to about the topic. There may be better info out there but I couldn't find it

Other guidelines I planned to follow
  • Use shingle below and above waste pipe (this comes from the regulations, and is very clear, although from speaking to plumbers they dont always follow the regs it seems, again I think a bit of common sense goes a long way, but at same time some people may be cutting corners to save time and cost)
  • Use conduits for everything (other than waste pipe which just sits directly in the shingle without a conduit). This conduit is probably not strictly necessary, and although it does add a little extra protection, the main reason for them was to allow for easier fixing of any disaster scenarios that may occur. Any of the pipework or cables could in theory be replaced entirely without the need for any digging. In theory :)

So here are two of the sketches I made when planning things out:

Trench design.jpg


Trench design 2.jpg


The final design didn't end up quite like that, partly because I added more separation between the power and data lines, and partly because the plumbers decided to go a bit freestyle jazz once they started work, and just did it their own way anyway.
Another modification was to have a layer of shingle both below and above the waste pipe. I ordered 4 tons of shingle, despite the plumber recommending only 1 ton - I got that wrong, but in my defence so did the plumbers as we used about 2.5 tons, so something in between would have been best. As a result of this I ended up with an unplanned gravel feature underneath my garden room.

Extract from the regs below shows how they stipulate a certain type of soil above the pipe, I decided that it would be easier to simply use more shingle rather than manually filter / select the soil, this guarantees no issues with stones / particle size.

waste pipe regulations.jpg


Drawings for the plumbers
Because I had quite a few complex discussions with the plumbers about how this would all be done, I wanted to document it to avoid any communication issues in the future. At the time it felt like complete overengineering, but this later turned out to be absolutely critical for getting us all on the same page both before work starts and also during (when I could see that said drawings were not being followed).
One example of this is when I noticed them only installing a single combined conduit for power and CAT6, and this is not what I wanted. After a discussion when they tried to convince me that a combined conduit was ok, I showed the drawings of what we agreed and they had no option but to add a second separate conduit for CAT6.

The overall design which I agreed with plumbers was to use a mixture of an open trench, which I would dig for any routes in the garden, and to use a 'moled' trench for the runs alongside the house. Not sure on the terminology here, but they use a machine called a mole to install the conduits underground, which they then feed the services into. The moling machine uses a small pilot trench for the entry and exit points, which they dig manually with spades, but it 'tunnels' its way underground using high pressure hydraulics without the need for an open trench. You cant install a waste pipe that way, its too big, therefore I couldn't avoid a large open trench in the garden

The pictures below show the plan: solid lines are open trench, dotted lines are underground conduits installed by the mole. Green is internet, red power, blue water, yellow waste. Black boxes are the mole entry points

trench top view 2.jpg


trench top view.jpg


trench view 4.jpg


trench view3.jpg



trench view5.jpg


Digging the Trench
I dug the open trench myself to save cost, and also because I wanted to learn and be able to do as much of the build as I could. I did consider doing the plumbing / drainage part also, but decided it best to see how the pro's do it, maybe next time I would try doing a bit more of the plumbing.

Again, the tools you select are very important here, and it is the same situation as with the initial groundworks in terms of which tools are the best. The mattock and the round point shovel were the winners for this job. I thought the normal square shaped shovel (with the cupped sides) would be really effective here once the mattock had loosened the ground, but that wasn't the case at all, it was the round point shovel that was able to extract the loose dirt from the trench once it was dislodged because the square one just couldn't penetrate the soil once it was loosened into big chunks.

Manual digging versus mini digger
From memory I think the trench took me about 3 days to dig, it is 800mm deep, 400mm wide and about 12 metres long. Not sure how long it would take with a mini digger, maybe just a day, so that may or may not make sense for you instead of manual spade work, depending on hire costs and how much you value your own time etc.

I will put some pictures of the trench digging in the next post
 
Marking out the sides with string
IMG_20210415_150539.jpg


Marking out the placement of the waste pipe for the toilet
IMG_20210419_112924.jpg


Down to depth for the first section of the trench
IMG_20210417_165414.jpg


second section of trench now complete - the kink was deliberate as i wanted to keep the trench the maximum distance away from the groundscrews.
IMG_20210508_104351.jpg


Mole evidence (you can see two of the three pilot holes in this picture, the shingle is also in place above the waste pipe)
IMG_20210508_135719.jpg


The other mole entry / pilot hole (this one is for internet)
IMG_20210508_143926.jpg


The mole machine
IMG_20210508_143934.jpg


Internet (CAT6 x2) and power (SWA)
IMG_20210522_161518.jpg


Same view from the future
SWA and CAT6.jpg


Water pipe within blue water pipe conduit which is also inside grey waste pipe conduit
IMG_20210522_173138.jpg


Foam insulation gets added later to the water pipe to stop it freezing
I am interested in knowing how people handle this type of situation with water pipes emerging from the ground in and around waste, what I did seemed pretty strange but I couldn't think of a better way of achieving what I needed.
 
Some additional trench pictures showing some other details:

Skip needed for removal of dirt - this ended up about 80% full just from dirt
IMG_20210528_151759.jpg


The SWA cable was just pushed through the plastic conduit from one end and then popped out at the other end. The CAT6 was pulled through after first pushing through that yellow thing you can see. The yellow thing probably has a name (maybe a pull rod or something like that) and I borrowed it from a BT engineer that just so happened to be working on my house at the exact same time as I was doing the conduit stuff.
IMG_20210519_141405.jpg


Ever wondered what 4 tons of pea shingle looks like?
IMG_20210508_095703.jpg


Me making curved piping because I didn't want to go and buy some
IMG_20210522_154713.jpg


The inspection chamber
IMG_20210417_171924.jpg

IMG_20210417_171910.jpg


IMG_20210417_171850.jpg


The plastic conduit for the water pipe needed trimming down - easier said than done
IMG_20210522_110914.jpg


IMG_20210522_164546.jpg


I was a bit disappointed in how the plumbers had left this part for me, and how much work was required to finish it off. The conduit was way too stiff for me to be able to bend it into shape so that the water pipe came up under the building next to the waste pipe. So I needed to trim down the conduit so that I could bend the water pipe into position and then install it into a suitable 'pre-curved' piece of conduit. Trimming down the outer conduit with the multi tool was quite risky / tricky with the water pipe is already inside it but I got it done without accidentally cutting into the inner pipe.
An additional conduit was eventually added at the end and is made out of grey waste pipe. I custom made its shape using the mitre saw, super glue and gorilla tape. It has multiple compound angles in it so I couldn't just install a 90 degree corner piece. This didn't feel like something you would see a professional doing but I didn't have any better options I could think of.
 
May 2021 - Bi-Fold Threshold Decision

So around the same timeframe as doing the trench I also needed to decide what threshold I needed for my doors, which at this stage had just been kicked off.
I had a really hard time deciding on this one, because I wanted a low threshold bi-fold for a seemless transition between inside and outside, but at the same time I needed it to be weather tight (obviously) and hence perhaps needed to go with a rebate threshold.

I knew nothing about this stuff previously so I spent a fair bit of time looking into the pros and cons. I did a forum post about it actually and received some good advice.

This picture should give you an idea of what I'm talking about, a threshold is the bit at the bottom of the door above the cill where it meets the outside world - it is the thing that stops or allows things getting in and seals the building from the outside world.
Bifolding-Door-Standard-Threshold-no-cill.png


The two types of threshold
For those that don't know (like I didnt until recently) there are two types of door threshold:
  1. Rebated threshold - fully weather proof, but has a small lip at the rear side of the door to stop rain getting in
  2. Low profile threshold - not as weather proof as rebated but doesn't have the lip at the rear meaning you have more chance of creating a seemless transition from outside to inside
See the drawings below which I managed to get from the manufacturers (note that the cill is not actually shown in the drawing).

Low profile threshold
low threshold.jpg


Fully weathered / rebated threshold
standard threshold.jpg


The door suppliers typically play it safe and insist that you need a rebated threshold if the door is 'outside'. However, it is not quite a clean cut as that, because in reality, if your door is outside but is under a large canopy, it could mean that rain doesn't ever reach the door, or you may have a small canopy and some rain could reach the door, but not as much as if the door was unprotected.
Also, when you look at the design of the low profile threshold, it is actually pretty difficult to imagine water getting in, my reasons as follows:

  • There are two sets of brushes, these should stop drafts but also some of the rain I would imagine
  • There is a drip channel, meaning that if water does reach the cill / lower rail, it will simply drain out again
  • There is a rubber seal on the outside to prevent water even getting to the cill / lower rail area (Note: I didn't realise it had this rubber seal until after the threshold decision had been made, because it isn't on the drawing)
Lucky accident
Anyway, what happened is I couldn't decide so played it safe and ordered the rebated threshold. As it turned out the door company screwed up the order and delivered me a low profile threshold for both sets of doors (bifolds and french doors). This turned out to be a bit of a winner because I negotiated a £1000 discount, and at the same time I sort of got what I really wanted all along which was a low profile threshold - I think this will be much nicer if I can have a single level from inside to outside without the lip. Now that future Martin has seen the doors in the flesh I feel confident that this will be weatherproof, especially as I have a small canopy, but time will tell.

Ultimately, a low threshold is a compromised design because it relies on the little brushes and the little rubber seal trim to protect you from the rain, rather than a proper rebated lip which could be more robust over the years to come. But I will find out if this turns out to be a good 'decision' or not after a few years.

Cill size decision
Is it cill or sill? I see both spellings on line. Anyway...
One other decision to make at this stage is the length of the cill - there were two options with my door manufacturer, 90mm 'stubby' cill and 150 'normal' cill.
I ended up chosing the normal 150mm cill which turned out to be the right decision. Once you add the battens and cladding to the outside of the building, a stubby cill would actually be buried behind it! The 150mm was in fact only just long enough, but this is partly to do with the placement of my door / cill (which I had placed in the middle of the 5 x 2 stud walls). If you needed longer effective length on your cill to clear your cladding, you could position the cill and door further forward inside the wall - this placement seems to be a fairly arbitrary decision and you just get asked when the door installers arrive on site what you want.

I have horizontal cladding therefore a single layer of battens, but if you go for vertical cladding then you will need two layers of batten. This adds 25mm and could mean the 150mm cill was too short, unless you positioned the door a little further forward. Some of this stuff needs thinking about at this stage, hence why I am mentioning it.



Martin
 
You were right in using the pea gravel,it serves to protect the pipework/cables by providing a cushioning effect against ground heave/movement should there be any, and in years to come when you have won the lottery and moved into a mansion;);)the new owner when digging a hole for a pond , will first hit the gravel which should alert them to the hidden stuff before doing any damage. Will you be fitting a drainage channel as shown in the drawing for a low profile cill ?.
 
Hi Mark, no I'm not planning on adding the drainage channel as shown on the low-pro drawing, I don't think I need it. I did spend some time thinking about adding an ACO drain along the front and looked at the options, but I decided that it isn't necessary. I will have composite decking in front of the building and this has gaps between the boards which is where the water will drain off. I do want to keep water away from the building of course, and I have ensured that I have a small gap in between the decking and the building to prevent water running towards the building. The decking is also angled at 1 in 80 gradient away from the building. The bearers for the decking is normal tanalised C24 but I'm planning on painting the tops of it in bitumen paint.
 
May 2021 - Wood Delivery

Towards the end of May I received my main materials delivery
This was pretty much everything needed for the base/floor, walls, roof including wall and floor insulation (but not roof insulation, that would arrive later)

IMG_20210525_114935.jpg


IMG_20210525_115242.jpg


Mistakes

Mistake 1
I made a big mistake at this stage because I stupidly ordered it all together rather than staging the deliveries for when I actually needed the materials. I was lucky because I called the timber yard and they agreed to take back some of the stuff on the lorry and deliver it later when I needed it. For example battens, roof boards, plasterboard. Not only did I not really have room for everything in the garden but the plasterboard would probably have been damaged if stored outside, even if covered as I am sure it would have got wet at some point.

Mistake 2
One quite minor mistake I made was not really preparing for the delivery, I had cleared a space and erected a gazebo but hadn't put any bearers down on the ground to support it all, so that was done in a last minute panic when the lorry arrived, which wasn't much fun. I was lucky I had some wood I could use, otherwise I am not sure what I would have done.

Mistake 3
I didn't think about the build order and hence how I wanted to store everything. It is no trivial matter to get to a few lengths of wood which are stored at the bottom of a pile, handling wood multiple times, just to get access to things underneath it is a major time drain and something I would plan for much better next time. I may also simplify my wood length configurations - I actually created a little excel sheet when I was doing my wood ordering calculations, it took all the lengths which I needed and told me what lengths I should buy based on maximum utilisation and the price of each length (yes some lengths are cheaper than others per metre for some reason). For example, if you need 50 joists at 1.1m long you might think buying 2.4m lengths is pretty good going in terms of waste, until you realise you can do better with 4.5m lengths and save quite a bit of cash. Also, sometimes its worth buying a length that appears to give more waste, because if the offcut is exactly the same size as your noggin length, then that ends up being zero waste. So the spreadsheet worked all that sort of stuff out and hence I ended up with virtually zero waste but at the same time a stupidly complicated timber configuration (I think I had 10 or 15 different types of CLS timber). In future I would simplify a little. Also, big long lengths like 6m are quite annoying to handle, especially if your mitre saw stand is quite basic like mine and doesn't support long lengths very well, so all that stuff would be a future consideration and I expect I will simplify.

Screenshot of my basic timber optimiser

timber optimiser.jpg


Mistake 4
The wood did get wet, because some of the timber lengths were 6m long and were longer than the gazebo. This meant the water ended up running into the main pile and getting a lot of the stuff at the bottom wet. Also, I didnt put the back on the gazebo and the water was able to get in sideways - not a lot but it built up over time. Because I have so much sat wood in a pile for quite a few weeks, the stuff at the bottom got a bit 'secretly moist' and ended up with a bit of mould on by the time I came to use it. So I ended up drying that all out in a makeshift rack and sanding it down once dry to get rid of any mould. I also bought a moisture meter to make sure it was all below 19% before I used it (I think that was the number I found on line for the moisture threshold). As it turned out it was mostly between 10 and 12% so it was fine and only took about 2 days to dry out.

Mouldy wood
IMG_20210624_183355.jpg


Drying racks
IMG_20210624_184134.jpg


Martin
 

Attachments

  • IMG_20210525_115230.jpg
    IMG_20210525_115230.jpg
    147.6 KB · Views: 5
Absolutely fascinating, astounding attention to detail and clear reasoning for your choices. Thank you for posting it.

Would a plumber's pipe cutter have done the job here, with less risk?".......Trimming down the outer conduit with the multi tool was quite risky / tricky with the water pipe is already inside it but I got it done without accidentally cutting into the inner pipe......"

I hope you've planned for a fridge in the garden room. You'll deserve a cold beer or two when this is finished!

Duncan
 
Absolutely fascinating, astounding attention to detail and clear reasoning for your choices. Thank you for posting it.

Would a plumber's pipe cutter have done the job here, with less risk?".......Trimming down the outer conduit with the multi tool was quite risky / tricky with the water pipe is already inside it but I got it done without accidentally cutting into the inner pipe......"

I hope you've planned for a fridge in the garden room. You'll deserve a cold beer or two when this is finished!

Duncan

thank you Duncan! glad you are enjoying it :) plumbers pipe cutter is a great idea, unfortunately the water pipe had a large valve on the end and the outer conduit was not wide enough to slide over it, so i had to cut along the length of the conduit to get it off. I am starting to think i need a beer fridge in there, it's been hard work to be honest....
 
How is your future self holding up?
:ROFLMAO:
future self is feeling a bit worn down and has resorted to writing this build log in order to avoid doing the work he is supposed to be doing. At the moment my activity splits most days seem to be about 20% building, 80% researching, head scratching and decision making....
 
:ROFLMAO:
future self is feeling a bit worn down and has resorted to writing this build log in order to avoid doing the work he is supposed to be doing. At the moment my activity splits most days seem to be about 20% building, 80% researching, head scratching and decision making....
keep up the well thought out work...whilst ignoring your future self's hindsight.
 
Jun 2021 - Base (Timber Frame)

I was pretty pumped at this stage because it marked (in my mind) the real start of the build; everything else up until this point felt like preparation so it was good to finally do some work that felt like I was creating something instead of just pushing dirt around.

It was a very quick job to roll out the weed membrane and cut a few holes for the groundscrews to poke through. I just made 'X' shapes with a stanley knife and it worked perfectly. You can buy ground staples to fasten down the weed membrane but that didn't feel like it was needed here so I didn't bother.

IMG_20210529_111511.jpg


Timber Base
I had CAD to work from, so I knew exactly what I was doing (lol). Even so that didn't stop me spending most of the day walking around and staring at the ground and figuring out where to start. Despite all that thinking time, I still screwed it up, almost straight out of the blocks (which was funny until I discovered just how difficult it is to remove 90mm nails from CLS). See if you can spot the mistake in the below base design.

original base design.jpg


Mistake 1 - no access for nailing
I concentrated on getting the perimeter in place first (correct dimensions and making it square), this felt like a good start (but it wasn't). The plan was to then build up the centre of the frame with additional bearers and joists once the perimeter was built because then I would know its the right size and shape. Unfortunately I didn't really think about how I was going to get nails into the joists, through the sides of the bearers. The perimeter is made up of doubled up 4 x 2, which gives a total bearer dimension of a 4 x 4, which is about 90 x 90mm. My framing nails are 90mm which means they will not pass through the 90mm bearers into the joists. So what I think I should have done was build the perimeter with single 4 x 2, and then makes 3 sub frames (I call these cassettes but that might be made up terminology) and then slot these 3 sub frames into the perimeter frame. That would effectively give me a doubled up 4 x 2 around the perimeter once it was all together. You may need to look at the CAD screenshot to know what I'm talking about there.

Updated Base Design
What I have done with the CAD below is update the left hand side with the new design (orange and yellow) and the right hand side I have left as the original design (blue).

new base design alongside original.jpg


New Design
In the new design the orange perimeter would be constructed first, to get the perfect size (but not shape at this stage) and then the yellow subframes would be constructed and slotted into place. Once the yellow subframes are constructed, I would then make it square by measuring the diagonals and 'tapping' it into shape. Once both diagonals were identical I would throw a long timber down (diagonally) and fasten in place with screws to keep it perfectly square.

new base design.jpg


Old Design
In the old design, what I actually did was construct the blue perimeter first, with doubled up 4 x 2 for bearers, only then to realise that wasn't going to work because I couldn't fasten the joists in place. I don't think I actually fastened it ALL together before I realised but I certainly threw a few nails into some of the frame as I remember having to pull them apart.
Note: you could I suppose proceed with the blue design and use joist hangers only to fasten the joists in but
  1. I didn't know if that would be strong enough
  2. I wasn't sure how easy it would be to fasten joists in place using only hangers, without first fixing from the side with the nail gun.
old base design in blue.jpg


Here is me implementing mistake #1:
IMG_20210529_132332.jpg


Mistake 2 - groundscrew placement irregularity
The groundscrews do not always end up millimetre accurate in terms of their placement (they hit stones and harder ground and move around a little bit as they are going in). For this reason it would make a lot of sense to first (before building the base) measure the final placement of every groundscrew so you know what you are working with. I am estimating here, but I would say that the placement of each screw is accurate to within 10mm or so on average - a lot of them were within 5 or 10mm but the odd one was a little out by maybe 10 or 20mm. Doesn't sound like a huge amount but it does affect things, especially when two screws are out in opposite directions thus creating a compound error.
The final placements of the groundscrews have a dependency to the size and location of your base.

Size of the base
Ideally, you want the size of the frame to be exactly as you designed it (in my case 9m x 3.7m) but this might not be possible depending on how the groundscrews fall. In my case I didn't really think this through and hence built the frame before thinking about any of this. Fortunately it was fine, other than having to grind down the sides of one of the groundscrews becuase it was sticking out from the base a little even though all the other screws were perfectly in line with one another.
Location of the base
You can shift the base around on the groundscrews to find the best compromise in terms of the location - you want as much of the bearer as possible to sit on the groundscrew and you want as little of the groundscrew overhanging away from the bearers as possible. So it's a case of messing around a bit to find the best placement.
Note: this is significantly easier (so I imagine) if you first build the perimeter and shift that into position rather than trying to move the entire base once its all built. I think I ended up using a club hammer to 'tap' it into position until I was satisfied (or until I got bored)
Note: it is critical that you have a cross brace on at this stage so you can manipulate the placement of the frame on top of the groundscrews and keep the frame perfectly square.
Note: one thing that future self has been telling me about is that any braces, for keeping things square or plumb, should be kept in place for as long as possible and should never be removed prematurely unless they absolutely have to be, even if you think other bits of the structure are keeping things square or plumb, they may not be. Future self is talking about that quite a lot, so I guess its important.

Mistake 3 - I need a base to build the base
This was a bit of chicken and egg, I needed a large flat surface to build the base. Ironically, once the base is built I have a perfectly flat surface of the exact same dimensions I need to build a base. Anyway, this wasn't such a problem in the end, I simply placed long timbers across a few of the groundscrews in order to give me the support I needed.
Note: if you have two people this would be a bit easier because you wouldn't need the support timbers; each person could simply hold the joist in position with one hand and then nail it into the bearer with the other hand (assuming a nail gun). Doing this on my own I needed a way to have the joist supported at both ends before I could nail it. Later on I discovered that I could use the already built subframes in order to build more subframes as each subframe was almost exactly the same size. So this worked fine, or at least it did once I had realised I could do it that way.

Joist hangers
Not sure if you need nails from the sides into the joists AND joist hangers, but that's what I did. I imagine joist hangers on their own would be enough but I didn't know so did both. Anybody know the answer to this?
In terms of assembly sequence, some people attach hangers to the joists then attach that 'joist / hanger' assembly to the bearer. Other people nail the joist to the bearer first and then attach the joist hangers. I personally attached the hangers to the joists first because I thought it would be faster in the long run. I made 10 or 15 joist / hanger assemblies, and then made a sub frame by attaching them to the bearers.

Joist hanger assembly station.;
IMG_20210530_105859.jpg


Note that I made a little seat and a sort of wooden 'anvil' too because I found that a solid surface to hammer on made a huge difference to the speed at which the nails went in.

Stop Blocks for repeated cuts
I had quite a lot of joists to cut at the same length so I created a simple jig to allow me to do that on my mitre saw.
IMG_20210529_145130.jpg

IMG_20210529_145135.jpg

Caution: I would be careful if using a system like this; the timber base plate which I screwed onto the saw is what holds everything in place and keeps your cuts repeatable. I only used one screw on each side of the blade and although this felt quite solid, I think repeatedly banging long lengths of timber up against the stop block acted as a battering ram and moved the whole base plate by a mm or two. This meant that the joists were getting bigger and bigger each time I cut one until the final ones were maybe one or two mm bigger
Note: My basic evolution mitre saw stand is cheap yet basic, and the in built stop block facility on it is almost useless because the range on the extending arms is quite small. Defo not criticising evolution for this because the stand is only about £60 and I think its good value but in hindsight I should have bought the Dewalt one at about £150 which has a far bigger range, and is just a bit nicer from what I can see

400mm spacing (not 400mm centres)
PIR is expensive and hence I wanted to maximise utilisation / minimise the waste. I am using 22mm floor boards, which fit together with a toungue and grove, and are glued together with a strong PU glue. For this reason I feel that the board edges don't need to fall on joists which gives me the freedom to place my joists wherever I want them. I made a 400mm spacer jig to help me get the spacing between joists accurate.

IMG_20210530_120058.jpg


I will include a few more pictures in the next post as I have reached my 10 limit for this one :ROFLMAO:

Martin
 
In the below picture you can see the subframes. For some reason I am levering up one of the subframes, no idea why I was doing that, possibly to put more nails in, I don't know.

IMG_20210530_144430.jpg


The below picture is a good representation of how I was constructing the subframes by the end. Note that this still isn't how I would do it in the future, because in the below you can see that I wasn't adding joists to the end of the run, instead I was utilising the longer perimeter timber to terminate the run (if that makes sense). This turned out to be problematic because in some areas I couldnt get access to punch nails in from the sides. Yes that's right, the exact same mistake from earlier. So if you look at the orange and yellow CAD from earlier, my improved design, you will see that each subframe has its own end joist, which could then be fastened to the perimeter from the inside. I would be surprised if anybody is following what I am saying here....
IMG_20210601_101824.jpg


Frame complete and cross braces in place:
IMG_20210609_104315.jpg


Attaching the frame to the groundscrews:
IMG_20210611_150306.jpg

Note: to the left of the bracket (and also sticking out of the bracket) you can see the coach screws which I used to bolt everything together. I put in about 50 of these to tie it all the bearers and subframes together. I also used a forstner bit to create a rebate for the head of the bolt, so that I had a clean aperture for the PIR. It might be quicker to cut a tiny slot in the PIR in the future...

Coach screws going in (I asked my wife to do this, just for fun - I think she enjoyed it)
IMG_20210611_153631.jpg


Trimming one of the groundscrews:
IMG_20210622_105845.jpg


IMG_20210622_105851.jpg


you can ignore the floor boards and PIR, those are from the future and will be covered in the next post.
that's all for now

thanks
Martin
 
Your calculations on beam deflection above the doors may or not consider:

The structure gets built and roof goes on before doors go in….therefore the dead load deflection will occur before doors go in

door frame is presumably a bit shorter than structural opening….so when you pack the frame, the dead load deflection does not impact on the door frame.

obviously live load, eg snow will impact.
 
Back
Top