An enclosure is secure when five questions can be answered about it, and the breaking strength of the mesh is not one of them. A panel rated far beyond anything its animal can generate will still admit a neonate through a 100 mm opening, still fail where it meets a galvanized post, and still hide a broken strand behind a curtain of ivy.
The five tests below decide the outcome. They run in an afternoon, need no instruments, and apply equally to a welded panel, a hand woven mesh, a glass wall or a line of bars. They measure not strength but whether the barrier is still a barrier everywhere it is supposed to be.
Executive Summary
Security belongs to the whole enclosure envelope, not to the panel. This guide sets out five tests a zoo can run in one walk round: what the aperture actually admits, where the envelope stops being continuous, what has been added since handover, whether the surface is still the alloy specified, and whether a defect would be visible before it became a hole. It also covers what to record, and when a panel should be replaced rather than re-tensioned. The reference specification is the animal enclosure mesh product page.
Send us your enclosure layout for a panel schedule
Quick Answer:
What makes an enclosure secure is continuity: an aperture small enough to exclude the smallest thing that must not pass, a base and top detail that close the envelope against ground and roof, and a material that is still the grade specified five years earlier. Strength decides whether the barrier survives the first bad night. Continuity decides whether there is a way through on the ten thousandth ordinary one.
Key Takeaways
- The aperture is set by the smallest animal that must not pass, not the largest one being held.
- Most security is lost at four junctions, not in the middle of a panel.
- Every load added after handover, from a hung feeder to a climbing vine, is a point load the panel never saw.
- A 304 panel near salt water will stain, and the stain is the earliest readable evidence that the section is going.
- A defect you cannot see from the path is not a defect you have, until the day you do.
Why security is a property of the envelope
A security review usually starts with the animal and ends with a strength figure. That order is backwards. Start with the boundary and ask where it is breached in practice. The answer is almost never the middle of a 9 m by 18 m panel, which is the part everyone photographs and the part carrying the largest safety factor. At site scale the same inversion applies, and the enclosure solutions overview separates enclosure classes before any panel is specified.
Anyone who wants to know how to test whether an enclosure is secure can begin with the wording regulators already use. Under 9 CFR 3.75, primate housing must be “structurally sound for the species” it houses, “kept in good repair”, and must “protect the animals from injury, contain the animals securely, and restrict other animals from entering”. Five obligations sit in that sentence, and a purchase order usually addresses only the first. Sound structure is a specification. The other four are properties of how the envelope was detailed and how it is maintained.
That distinction is what the five tests measure. They are a walk round, not a calculation, ordered by how often each one turns out to be the cause.
Test 1: What an aperture lets through
The first question is not whether the animal can escape. It is what an aperture lets through, because that is what catches the failures small institutions actually suffer: a neonate that slips a 50 mm opening, a keeper’s hand that reaches a latch through a 60 mm one, a rat that treats 25 mm as a doorway, a bottle cap posted through a gap that looked harmless on the drawing.
Set the mesh aperture from the smallest body dimension that must be excluded, then check it against every age class the enclosure will hold. Stainless steel rope mesh can be specified anywhere between 20 mm and 102 mm, and the difference between 50 mm and 76 mm is not cosmetic: it changes what the panel stops, what falls through, and how much of the animal’s view is wire. Rope diameter matters as much. A 1.5 mm rope at 50 mm reads as nearly open; a 3.2 mm rope at the same pitch reads as a grid and behaves differently when something heavy leans on it. For small primates and parrots the common answer is a 1 in × 1 in opening, which is HM1625 in 1/16 in wire.
Then look at what sits inside reach that is not the mesh: fixing heads, exposed bolt thread, a tensioning fitting, the top rail. A panel with a 40 mm aperture and a 90 mm fixing head is a 90 mm aperture with extra steps.
Test 2: Where the envelope stops being continuous
Most of the time, where an enclosure loses its security is not a place anyone drew. It is a transition, and four of them matter.
The base comes first. How the base detail is sealed against digging decides whether the envelope ends at the ground or continues below it. A panel that stops at grade is a fence; a panel that turns and runs below grade, or meets a concrete curb with no gap, is a containment boundary.
The top comes second. An open-topped yard is only secure against species that cannot climb or jump it, and the rail carrying the mesh is usually the weakest element in the upper third of the panel.
The post line is third, and it is where the envelope most often stops being continuous silently. A panel fixed to the face of a post leaves a vertical slot at every post; a panel fixed between posts does not. No strength calculation notices.
Penetrations are fourth: a drain, a water line, a cable duct, a service hatch, a camera bracket. Each is a hole cut through a boundary by somebody thinking about something else. Add them up on the plan before you walk, then go and look at each one. Excluding animals that are not supposed to be inside at all is a separate problem from containing the resident, and wildlife netting is specified on exclusion logic rather than containment logic.
Test 3: What gets added after handover
Every enclosure is commissioned with a clean panel and then lived in. What gets added to an enclosure after handover is the question no specification answers, because the additions are made by people solving a different problem on a Tuesday.
A shade cloth tied to the top rail adds wind load and a point load at each tie. A feeder hung from a mid-panel crossing puts sustained pull on two strands instead of a distributed one. A climbing rope, a hose left coiled on a rail, a sign bolted through the mesh, a camera bracket, a vine planted for screening that now weighs 40 kg wet: each is a load the panel never saw, and each is a corrosion site, because every new fixing is a new pair of dissimilar metals in a wet environment.
The test is a count. Walk the enclosure and tally everything touching the mesh that is not on the original drawing. Above zero, the panel is carrying something it was not specified for. The fix is to remove it or move the load onto the frame, where it belongs. What the frame and fixings should be made of is set out in enclosure materials for sensitive species, which treats grade, finish and fixing as one decision.
Test 4: Is the surface still the alloy you specified
Why the material you specified is not the material you got is a question answered by looking, not by reading the order. The panel in front of you is the only evidence that matters, and it has been reporting for years.
Start with the wire. How to read a rust bloom on the wire is worth teaching to whoever does the walk. On 304 stainless steel, a light brown bloom that returns to the same place each season is usually tea staining: airborne chlorides or carbon steel dust left from fabrication, sitting on a passive layer that never got a chance to re-form. It is cosmetic until it is not. On 316 stainless steel, which carries molybdenum for chloride resistance, the same bloom near a coast means the site is more aggressive than the grade assumes, and the answer is a wash-down schedule rather than a different panel.
Then look at what happens where two metals meet. A stainless panel bolted to a galvanized post, a zinc-plated bracket holding a 316 panel, a copper pipe touching a wire rope mesh: each is a small battery, and the less noble metal goes first. The corrosion resistance of the panel is not in question. The joint is. The wire itself is dealt with in animal fence wire that does not rust or sag, and the grade is only half the answer.
Then look at the cleaning regime. How cleaning chemicals shorten mesh life is rarely written down, because the chemical is chosen for the floor. Chlorinated cleaners, acid descalers and bleach left to dry attack the passive layer that makes stainless stainless, and the damage concentrates where dirt collects: under a fixing head, inside a ferrule, along a base rail. A panel washed with the wrong product twice a week fails sooner than an identical panel that was never washed.
Test 5: Would you see a defect before it becomes a hole
The last test is about evidence. How to inspect a barrier you cannot reach is the practical form of a harder question: how to keep an inspection sightline clear on a structure that grows and fills.
Three things block the sightline, and none is the barrier: vegetation planted for screening and never pruned from the mesh side; stored material along the back of a service corridor; and height, which puts the top of a 9 m panel out of view.
The fix is a line, not a ladder. Choose a viewing position for every panel, mark it on a plan, and keep it clear the way a fire exit is. Binoculars from a marked point resolve a broken wire at 12 m in seconds. A roof’s far face needs access, not inspection.
What a security walk round should record is narrow enough to repeat: panel identifier, date, what was seen, what was decided. A photograph of the same corner from the same position each quarter beats all of it: change is what you are looking for.
A broken wire is worth rehearsing. In hand woven mesh a broken strand is local and visible: the weave holds its neighbors and the break shows as a V. In a welded panel the equivalent failure is a weld that parted inside a joint that still looks closed. The first is a maintenance item, the second a hole with a lid on it. What the panel is made of decides how long either takes to appear, which is why the animal enclosure mesh range is specified by aperture and rope diameter, not strength.
What an enclosure has to prove to be called secure
Run the five together and they produce one standard, which is what an enclosure has to prove to be called secure. Every point on the envelope must be continuous, closable or visible. Continuous means the boundary turns and seals at base, top and every penetration. Closable means the aperture excludes the smallest thing that must not pass, including what visitors and keepers introduce. Visible means a defect can be seen from a marked position before it becomes a breach.
When a slack panel becomes a security problem is the moment its deflection opens a gap at a junction, not the moment it stops looking tight. Re-tensioning is right when the wire is sound and the slack is in the lacing, and wrong when the slack comes from a stretched strand, a pulled ferrule or a base that has moved. When to replace a panel rather than re tension it is decided by the base and the terminations, not by the middle of the sheet.
That is also where a specification stops being a shopping list: the tests exist to make the junctions hard to get wrong.
Ask us to review the five details on your drawing
FAQ
It raises the ceiling on what the barrier survives, and it does nothing for the four junctions, the aperture or the additions.
The smallest one any resident needs. If a 50 mm opening is required for a neonate or a small primate, the whole envelope takes 50 mm, including the panels around the larger animals.
No. It is a reason to clean, find the source of the contamination and check whether the grade suits the site. Replacement is the answer when pitting has taken the section.
Twice a year for a full walk, and after any event that moved something: a storm, a new enrichment item, a change of species, a contractor working on the boundary. The record matters more than the interval.