The ground under the building decides how much weather it can survive
Mid winter across southern Australia is the season that finds every weakness in an outbuilding. Cold fronts arrive off the Southern Ocean one after another, the soil never dries fully between them, and the equipment that spent summer under a tarpaulin is now standing in mud. A metal garage shed is the usual answer, and the wet months before spring are when most owners finally pour a base and put one up.
The temptation is to judge the purchase by the panels. The panels are the least of it.
A shell of this size weighs a few hundred kilograms. Spread across eighteen square metres of footprint, that is almost nothing, and a strong gust generates suction forces many times greater than the weight of the whole assembly.
So the structure cannot hold itself down. It borrows every kilogram of resistance from the base beneath it, which makes that base the first and most consequential engineering decision on the site.
A concrete slab performs three separate jobs at once, and it is worth separating them, because a shortcut taken on one of them is not a shortcut on the others.
- Mass: supplies the uplift resistance the shell lacks
- Plane: gives the base rail flat, square seating
- Barrier: keeps ground moisture away from the steel
Around one hundred millimetres of concrete over a compacted granular layer, reinforced with mesh and thickened around the perimeter, is the common specification for a footprint of this size. The thickened edge matters because the edge is where the anchors go.
Placement on the block is decided by water, access and the fire season
The position of the building is usually chosen for convenience and then regretted for drainage. The two considerations rarely point in the same direction.
The first question is where water goes during heavy rain. Every block has an overland flow path, and it announces itself only during the sort of downpour that arrives once or twice a year.
Placing a slab across that path turns the building into a small dam. Water banks up against the uphill side, which is precisely the condition that corrodes the base of a wall and heaves reactive clay under one edge of the concrete.
Access is the second question and the one most often underestimated. A vehicle needs a turning path in front of the doors, not merely an opening wide enough to drive through.
Six metres of clear apron in front of a three metre doorway is a comfortable minimum. Less than that means approaching at an angle and reversing in every single time, which is exactly the friction that leads to equipment being left outside.
Regulatory setbacks come third. Most councils specify minimum distances to boundaries and prohibit building over easements, and a sewer or stormwater easement crossing a block is not negotiable at any price.
In bushfire prone areas siting becomes a safety question rather than a convenience one. Separation from vegetation, and attention to the gaps where embers can enter beneath the eaves or under a door, matter far more than the material of the walls.
What sits under the concrete matters as much as the concrete
A slab is only ever as good as the material it bears on, and all of that preparation becomes invisible within an hour of the pour.
Topsoil is stripped because it contains organic matter that keeps decomposing and compressing. Building on it guarantees settlement, and settlement under one corner of a rigid slab produces a crack rather than a gentle tilt.
The replacement is a compacted granular layer placed in thin lifts and rolled between them, rather than tipped in to full depth in one go. Loose material dumped at full thickness compacts near the surface and stays soft underneath, where nobody can see it.
Winter complicates the work. Saturated subgrade cannot be compacted properly, because water sitting in the voids carries the load instead of the soil grains and springs back the moment the plate lifts.
A vapour barrier beneath the slab costs very little and settles an argument that would otherwise run for decades. Concrete is porous, and a slab in direct contact with damp ground draws moisture upward continuously by capillary action.
That matters to the steel sitting on it. The base rail and the bottom edge of every wall panel meet the concrete surface, and a slab that is permanently damp keeps that contact line permanently damp too.
Anchors fail in ways that have nothing to do with their strength rating
An anchor bolt carries a published pull-out figure, and that number is easy to trust. It is also conditional on details that get compromised on real sites more often than not.
Edge distance is the first of them. Concrete does not fail at the bolt itself; it fails by lifting a cone of material around it, and a bolt set too close to the perimeter simply breaks a wedge out of the slab edge.
Embedment depth is the second. The cone of concrete grows with depth, so a shallow hole drilled because the masonry bit struck reinforcement delivers a fraction of the rated capacity while looking identical from above.
Quantity is the third factor, and the most frequently ignored. Filling every second hole halves the restraint around the perimeter and doubles the load carried by each remaining bolt. The building looks finished either way, and the difference appears once in a decade.
Anchors also work in two directions at the same time. Uplift pulls them out along their axis while wind pressure on a wall tries to slide the whole shell sideways and loads them across it, and those are different failure modes acting on the same piece of steel.
Chemical anchors, in which a resin bonds a threaded rod into the drilled hole, distribute load along the full embedded length instead of wedging outward against the sides. They tolerate closer edge distances and repay the extra step on a thin or elderly slab.
The northern cyclonic coast and the southern temperate belt ask for two different buildings
Australian wind design divides the continent into regions, and the divide is not simply a matter of degree. It changes the character of the load rather than only its magnitude.
The southern and inland region covers most of the population, from Perth across to Adelaide, Melbourne, Sydney and the tablelands. Design there addresses strong frontal winds that arrive from a fairly predictable quarter and pass within hours.
North of roughly the Tropic of Capricorn, along the coastline near Broome, Darwin, Cairns and Townsville, cyclonic conditions apply. Design speeds are higher, but the more punishing difference is duration combined with reversal.
A cyclone works a building for many hours, and when the eye passes overhead the wind returns from the opposite direction. Every connection that was in tension goes into compression and back again, thousands of times over.
That repetition explains why cyclonic specifications call for more fasteners, larger load-spreading washers and heavier anchors even where the peak gust is only moderately higher. Metal that survives one extreme load can still tear out under many moderate ones.
Two site factors modify the regional figure, and both lie within the control of the owner.
- Terrain is one. A shell standing alone in an open paddock meets the full flow, whereas the same shell behind a house, a windbreak of trees or a neighbouring building sits in disturbed, slower moving air.
- Topography is the other. Wind accelerates as it climbs a ridge, so a hilltop or the edge of an escarpment can raise local speeds well above the figure that applies to flat ground a few hundred metres away.
Load walks a single path from the roof sheet to the concrete, and the weakest step sets the limit
Every part of an enclosure is a link in one chain. Wind lands on the outer skin, and from that moment the force has only one route available to it.
It passes from the sheet into the screws, from the screws into the purlins and girts, from those into the rafters and columns, from the columns into their base plates, from the plates into the anchors, and finally into the slab and the soil beneath.
The capacity of the whole assembly equals the capacity of its least adequate step. Doubling the thickness of the wall sheeting on a building held down by four bolts changes precisely nothing.
The joint between column and rafter is where the design earns its keep. If that corner were pinned, the frame would fold sideways like a parallelogram under any horizontal push at all.
Instead the knee is made rigid, usually by a bracket and several bolts holding the two members at a fixed angle. The corner then carries bending, and the frame resists sway through its own stiffness rather than depending on diagonal bracing across a doorway that has to stay clear.
The difference is easy to feel in your own body. A straight arm held loose at the elbow collapses when pushed, while the same arm with the elbow locked transmits that push into the shoulder.
A metal garage shed uses locked elbows at both knees of every frame, which is why the interior can be left free of posts and diagonal braces across the working area.
Column bases are the point at which a designed structure becomes a site-built one. The base plate has to bear flat on the concrete, because a plate resting on one corner concentrates the entire reaction into that corner and gradually crushes the material beneath it.
Purlins and girts do more work than their size suggests
Between the frames and the skin sits a secondary system of light members: purlins running along the roof and girts running along the walls, both spanning from frame to frame.
Their first job is to shorten the span of the sheeting. Deflection under a distributed load varies with the fourth power of the span, which is a far more dramatic relationship than intuition allows for.
Halve the distance between purlins and the sheet deflects roughly one sixteenth as far under identical pressure. Adding a purlin is therefore a much better answer to a flexing roof than buying heavier steel.
Their second job is less obvious and rather more important. A rafter under load has a compression edge that wants to buckle sideways, twisting the member out of its plane, and it will do so at a load well below the strength of the steel itself.
Purlins bolted at intervals along that edge hold it in line. The rafter cannot snake sideways because it is tied at short intervals to members running the other way, so the full bending strength of the section becomes available.
That explains why a partly assembled building is far more vulnerable than a finished one. Frames standing without their purlins, or purlins without their sheeting, have none of the restraint the design quietly assumes is present.
The finished skin then contributes something of its own. Fastened along all of its edges, a sheet resists being pushed out of square in the way a stretched panel resists shear, so the wall behaves as one continuous surface rather than as a row of independent posts.
Assembly order is a structural specification in disguise
The instructions supplied with a kit read like a sequence of convenience. They are usually a sequence of stability.
Members are stable in the completed configuration and unstable in nearly every intermediate one, so the order exists to keep the partly built structure standing while it is still incomplete.
Squaring comes before fixing. Measuring both diagonals of the base rectangle and adjusting until they agree is a five minute operation that determines whether the doors fit at the end of the day.
A rectangle ten millimetres out of square at the base propagates upward. The frames lean fractionally, the purlin holes stop lining up, and the accumulated error announces itself at the very last sheet, when nothing can be moved.
Bolts go in loose and stay loose until the geometry is right. Tightening progressively as the work advances locks in whatever error exists at that moment and makes later adjustment impossible without dismantling half the building.
Crew size matters more than the tool list. Two people can build one of these comfortably; one person cannot safely hold a wall panel in any wind at all, and a single panel behaves like a sail the moment it leaves the ground.
Every opening is a subtraction, and the double front doors are the largest one
A closed box is a remarkably efficient structure. Cut a hole in one face and the efficiency drops, not in proportion to the size of the hole, but faster than that.
Load that used to travel through the missing area has to divert around it, concentrating in the narrow strips of wall on either side and, above all, at the corners of the opening.
On this footprint the double front doors occupy most of one short wall. That end of the building has surrendered the greater part of its shear capacity, and the framing around the opening now carries what the wall used to carry.
There is a second effect, considerably more dangerous, and it appears only when the doors are standing open. An enclosed building and a partially open one experience completely different internal pressures.
With everything shut, air flowing over and around the shell creates suction on the roof and the leeward walls while the interior stays near ambient pressure. The roof is being pulled from the outside only.
Open the front doors to a wind blowing straight at them and the interior fills like a bag. Internal pressure now pushes the roof upward from below at the same moment that external suction lifts it from above.
The two effects add together. Net uplift on a building caught with its doors open can approach twice the figure the same building would see with them closed, which is why post-storm inspections so often find an intact structure beside a ruined one of the same make.
The operating rule that follows is unglamorous and free. Doors are shut and latched before a front arrives rather than after the first gust, and a door left open on a still winter morning is a decision that has to be revisited when the afternoon change comes through.
A header puts back the strength the doorway took away
The member spanning the top of the door opening performs structural work of a very specific kind, and its generous size is not decoration.
It is a bridge. Roof load and wind load that would have been carried by the missing wall are collected across the opening and delivered to the columns on either side, which are usually heavier than ordinary wall posts for exactly that reason.
Those jamb columns then pass a concentrated load into their base plates and anchors. The two anchors flanking the double doors typically see more load than any other pair around the perimeter, and they are the last two that should ever be skipped.
What actually governs the size of the header is generally not strength at all. It is deflection.
A beam can be entirely safe while sagging several millimetres, and a header that sags a few millimetres closes the clearance above the door leaves. The leaves rub, then bind, then refuse to latch, and the owner concludes that the hardware is faulty.
So the design limits deflection to a small fraction of the span in order to keep the doorway square under load. That is why the framing around the front opening of a metal garage shed looks heavier than the arithmetic of strength alone would seem to demand.
The doors themselves then need somewhere to rest when open. A hold-back catch fixed to the wall stops a leaf swinging into the wind, and a leaf that swings freely eventually tears its own hinge fixings out of the frame.
The side door changes habits rather than structure
A personnel door in the long wall is a small opening. Its header spans well under a metre, the shear it removes from that wall is minor, and the framing absorbs the loss without any special provision.
Its real effect is on behaviour, and behaviour is what determines how long a building lasts.
The double doors are the largest moving assembly on the property. Every cycle works the hinges, the latch bar and the bottom guides, and every opening admits whatever the weather happens to be doing at that moment.
In a wet July, opening a three metre doorway into a southerly means driven rain across everything stored inside and water on the floor that then has all night to sit there.
A side door reduces those cycles from several a day to a handful a season. Fetching a spanner no longer requires opening a wall.
- Daily access: side door, small opening, minimal exposure
- Vehicle access: front doors, opened deliberately, closed promptly
- Storm preparation: both sets latched, never propped open
Placement deserves thought before the panels go on. A side door on the sheltered wall, sitting on the natural walking line from the house, gets used; one facing the prevailing weather at the far end of the building quietly stops being used within a season.
Storms take buildings apart at the fasteners, not through the steel
Inspect the wreckage after a severe wind event and the pattern repeats itself. Sheets are found some distance away, largely intact, with a row of small torn holes along one edge.
The steel did not fail. The connection failed, and it failed by the sheet tearing around the fastener head.
The washer exists to answer exactly that. Spreading the clamping force over a disc many times the area of the head raises the load at which the sheet tears, which is why the larger cyclone washers appear in northern specifications.
Where the screw goes matters as well, and the two options trade against each other honestly.
A fastener through a raised rib keeps the hole above the water running down the surface, but it sits on a short lever above the purlin and permits a little flexing under load.
A fastener through the flat clamps the sheet directly against the purlin, which is considerably stiffer, but it places the hole in the path of running water and makes weather tightness entirely dependent on the sealing washer.
Driving torque goes wrong most often, and in both directions at once. Overdriven screws crush the sealing washer until it extrudes past the metal cup, and they dish the sheet around the hole into a small dimple that then collects water. Underdriven screws leave the washer uncompressed and the seal open to the weather.
Both look perfectly acceptable from ground level. The difference between them is a fraction of a second on the trigger of a driver, which is a strong argument for a depth-setting nose cone rather than for judgement.
Steel that expands twice a day slowly undoes its own fixings
A length of steel expands by roughly twelve micrometres for every metre of length and every degree of temperature rise. The numbers sound negligible until they are applied to a real building.
A dark roof sheet in the Riverina can swing fifty degrees Celsius between a clear winter night and the following afternoon under sun. Over six metres that is around three and a half millimetres of movement, out and back, every single day of the year.
The sheet is fixed at both ends. Something has to accommodate the difference, and what accommodates it is the interface between the screw shank and the hole it passes through.
Each cycle slides metal a fraction of a millimetre against that shank. Repeated a few thousand times, the hole elongates into a slot, the washer wears flat on one side, and the screw walks a small fraction of a turn out of the purlin.
This is why fastener tension is a maintenance item rather than a one-off installation task. Nothing has been installed badly; the building is doing what steel does in a climate with large daily swings.
It also explains why sheets are lapped rather than butted together, and why the fixings at a lap are arranged to let adjacent sheets move slightly relative to one another. A shell clamped rigidly at every point would tear itself apart trying to grow.
The maintenance implication is specific and saves time. Screws are inspected for backing out rather than for rust, and the fixings on the northern and western elevations, which take the most sun, come loose first.