Pet Carrier Bed: Comfy Rest

Pet carrier production desk · Updated 2026-10-06 · 16 min read

The difference between one dog carrier factory and another is rarely the stitch count - it is whether the pet carrier accessory pattern survives a 1.5x static load without permanent set.

Density, Indentation Force and What the Hand Reads

Three numbers describe a foam and the customer feels all three without knowing any of them. They are density, indentation force deflection and rebound resilience, and conflating them is the most common specification error in the category.

Density is mass per unit volume, 22-45 kilograms per cubic metre in this product. It is the durability variable and it correlates with price more tightly than any other figure. A foam at 22 kilograms per cubic metre costs 0.90-1.80 USD per kilogram and lasts 12-24 months of daily use; one at 38-45 costs 1.90-4.20 and lasts 36-72 months. Density is not softness and it is not firmness.

Indentation force deflection is firmness, measured as the force needed to compress a specimen to a stated percentage of its thickness. A soft bed is 60-120 newtons at 25% compression on a 323 square centimetre indentor; a medium is 120-220; a firm is 220-380. Two foams can have the same indentation force at very different densities, which is why a cheap bed can feel right in the shop and be flat in four months.

Rebound resilience is what the hand reads as quality. It is the percentage of energy returned by a falling ball, at 38-55% for a conventional polyurethane, 55-70% for a high-resilience grade, and 8-25% for a viscoelastic memory foam. A high-resilience foam springs back under the hand; a memory foam does not, and that difference is a design decision rather than a quality difference.

What the customer actually feels in the shop is the combination: a foam at 32 kilograms per cubic metre, 160 newtons indentation force and 62% rebound reads as a good bed. The same foam at 22 kilograms, 160 newtons and 42% rebound reads identically on the day and differently in month five, and the difference is entirely in the density.

Specifying by density alone is therefore the correct commercial discipline, and the practical minimum for a bed sold as an everyday product is 28 kilograms per cubic metre. Below that, thickness loss after 80,000 compression cycles rises above 12% and the bed visibly thins.

Thickness interacts with the three numbers. A 60 millimetre slab at 28 kilograms bottoms out under a 40 kilogram load; the same slab at 100 millimetres does not. Where the bed is thin, density and rebound have to rise together, and that is the specification for a travel bed rather than a house bed.

Specify density as the durability variable at 28-45 kilograms per cubic metre, indentation force as the comfort variable at 120-220 newtons, and rebound as the perceived-quality variable at 55-70%, and never substitute one for another.

Compression Set: the Only Figure That Predicts Flattening

Every complaint about a bed going flat is a compression-set complaint, and compression set is the one figure in the specification that is measurable, predictable and almost never specified.

The test compresses a specimen to 50% or 75% of its thickness, holds it at 70 degrees Celsius for 22 hours, releases it and measures the thickness not recovered after 30 minutes. The result is a percentage of the original thickness. A conventional polyurethane at 28 kilograms per cubic metre gives 8-15%; a good high-resilience grade gives 3-8%; a low-density foam at 20 kilograms gives 20-35%.

The practical threshold is 12%. A bed with a compression set under 12% recovers fully between uses and still looks new after two years; one at 20-35% loses a visible fraction of its loft permanently and the customer notices it between month four and month eight, which is exactly when a review is written.

Temperature accelerates it and the test temperature is chosen for that reason. At 23 degrees the same foam gives a figure 40-60% lower than at 70, which is why a supplier's room-temperature data sheet is not evidence. The specification has to name the test condition, and the standard method reference sits in the ASTM system.

Cyclic fatigue is a different mechanism and it is measured separately. A foam subjected to 80,000 compression cycles at 0.5-2 hertz loses 4-10% of its thickness at 32-38 kilograms per cubic metre and 12-25% at 22. It also loses indentation force, by 8-18% at the higher density and 22-40% at the lower, which means a cheap bed gets softer as well as thinner, and that is the combination that reads as worn out.

Humidity matters where the bed is used outdoors or in a damp room. A polyester-based polyurethane hydrolyses slowly above 60 degrees and 80% relative humidity, losing 10-25% of its tensile strength over 24 months; a polyether-based polyurethane loses 3-9% under the same conditions. Polyether is the specification for any bed that will be washed and dried repeatedly.

Recovery time is the customer-facing consequence of a low-resilience foam. A viscoelastic memory foam recovers its full thickness in 8-60 seconds after the load is removed, which is slow enough that a customer pressing it in a shop reads it as dead. That is a perception problem to be solved with a cover that hides the recovery, or with a hybrid construction of memory foam over a high-resilience base.

A compression set under 12% at 70 degrees for 22 hours is the only figure that predicts whether a bed stays flat-looking, and it has to be specified with the test condition named rather than taken from a room-temperature data sheet.

Memory Foam, Latex and Polyurethane Compared

Three foam families are in production for this product and the choice is a price-point decision, a durability decision and a shipping decision at the same time.

Conventional polyurethane slab is the volume material at 0.90-2.60 USD per kilogram. It is cut from a bun on a horizontal or vertical cutting machine to a tolerance of plus or minus 1.0-2.5 millimetres, it rebounds at 38-55%, and it is the lightest of the three at 22-38 kilograms per cubic metre. Its weakness is fatigue, and it is the material that generates the flattening complaint.

High-resilience polyurethane is the same chemistry with a different polyol and isocyanate balance, giving a more uniform cell structure. It costs 1.60-4.20 USD per kilogram, rebounds at 55-70%, and gives a compression set of 3-8% against 8-15%. It is 18-42% more expensive per kilogram and it is the correct specification for any bed above a 60 USD retail price.

Viscoelastic memory foam is a polyurethane with an open cell structure and a glass transition near room temperature, giving a rebound of 8-25% and a slow recovery of 8-60 seconds. It costs 2.40-6.80 USD per kilogram and it is sold on a pressure-relief claim. Its engineering weaknesses are temperature sensitivity, where its firmness rises 200-400% between 25 and 10 degrees, and breathability, where its low air flow of 0.5-3 cubic feet per minute makes it hot in summer.

Latex, natural or synthetic, is the premium material at 4.60-11.20 USD per kilogram. It rebounds at 68-82%, gives a compression set of 2-6%, and lasts 60-120 months. It is 2.4-4.6 times the cost of conventional polyurethane, it is 15-30% heavier at 60-85 kilograms per cubic metre, and it is the only one of the four that is a specified material for a bed sold on a ten-year claim.

Air flow is the specification that decides summer comfort and it is rarely measured. A conventional polyurethane passes 20-80 litres per minute through a standard specimen; a high-resilience grade passes 40-120; a viscoelastic foam passes 4-24. A bed that is returned as too hot is almost always a low-air-flow foam with a low-permeability cover, and both are measurable before the tool is cut.

Hybrid construction is the answer to most of these trade-offs. A 30-50 millimetre memory foam layer over a 60-100 millimetre high-resilience base gives the pressure-relief feel at the surface and the durability and support underneath, at a material cost 30-70% above conventional polyurethane and 40-60% below an all-memory-foam build.

High-resilience polyurethane at 1.60-4.20 USD per kilogram with 55-70% rebound and 3-8% compression set is the volume specification for a bed above 60 USD retail; memory foam needs a high-resilience base under it to avoid the temperature and air-flow problems.

Cover Fabric: Mass, Elastane Content and Abrasion

The cover is the part the customer touches and the part that fails first, and it is specified by mass, by stretch and by abrasion rather than by appearance.

Mass runs 220-380 grams per square metre in this category. Below 220 the fabric is see-through over a foam and it snags on a claw at 2-5 newtons; above 380 it is heavy, it dries slowly and it costs 1.20-3.40 USD more per bed for no perceived benefit. The working band for an everyday bed is 260-320.

Elastane content is what the specification note about stretch refers to. A woven or knitted cover with 4-8% elastane by mass stretches 15-35% in both directions and recovers 90-97% of it, which is what allows a cover to fit a foam tightly without wrinkling at the corners. Below 3% elastane the cover bagging at the corner is visible within 2-8 weeks; above 10% the fabric is difficult to sew without puckering and it costs 0.60-1.90 USD more per metre.

Recovery is measured separately from stretch and it is the figure that matters. A fabric at 30% stretch with 82% recovery is permanently loose after a month; one at 30% with 95% recovery is not. The test is a cyclic extension to a stated load for 100-500 cycles followed by a measurement of residual extension, and the acceptance is under 6%.

Abrasion is specified on the Martindale scale. A bed cover at 15,000 cycles pills and thins at the entry point within 3-9 months; one at 30,000-50,000 lasts 18-40 months. The entry point is where the animal turns before lying down, and it is the single highest-wear area on the product, which is why some programmes specify a heavier panel there at 380-460 grams per square metre.

Pilling is the cosmetic failure and it is measured on a scale of 1 to 5 after a stated number of revolutions. A polyester knit at 220-280 grams per square metre gives a pilling grade of 3-4 after 7,000 revolutions; a polyester-cotton blend gives 2-3. The specification is a minimum grade of 4 after 7,000 revolutions, and it is achieved by a tighter knit construction rather than by a higher mass.

Substance compliance on the cover is a certification question rather than a test question. A cover in contact with an animal for hours a day is specified to a textile standard, and the certificate that closes it is issued under the OEKO-TEX system, held per fabric lot and renewed annually.

A cover at 260-320 grams per square metre with 4-8% elastane, 30-50% stretch and recovery above 90%, 30,000-50,000 Martindale cycles and a pilling grade of 4, held under a current textile certificate, is the specification that survives both the wash and the claw.

Bolster, Base and Construction Geometry

The shape of a bed is a manufacturing decision disguised as a design decision, because every curve and every bolster is a cutting operation, a sewing operation and a foam yield question.

A flat mattress bed is the simplest construction: a rectangular foam block with a cover sewn on five sides or a cover with a zip. Foam yield from a bun is 82-94%, and the cover is two panels plus a boxing strip. It is the cheapest geometry and the one with the lowest sewing time at 3-7 minutes.

A bolster bed adds a raised perimeter of 80-200 millimetres height and 90-180 millimetres width, filled either with a solid foam ring cut from a bun or with a loose polyester fibre fill. A solid foam ring gives a defined shape and 40-90% higher material cost; a fibre fill gives a softer edge at 0.40-2.20 USD per bed and it migrates, flattening to 40-70% of its original height within 3-12 months.

Fibre fill migration is the failure to design against. A loose fill at 6-15 denier and 30-60 millimetres staple length migrates and clumps, and the fix is either a baffled channel at 120-250 millimetres spacing, at 0.35-1.10 USD, or a specified fill mass per unit volume of 12-22 kilograms per cubic metre that leaves no room to settle.

A cuddler or a donut bed with a continuous raised rim is a single moulded or cut ring with a centre well. Its foam yield from a bun is 55-75%, which is 15-30 points worse than a flat mattress, and that yield difference is the real cost of the shape. The alternative is a moulded part, which needs a tool at 4,500-18,000 USD and gives a yield of 92-98%.

The base is the component that is forgotten. A bed without a non-slip base slides on a hard floor at 15-40 newtons of horizontal push; the same bed with a printed silicone or a thermoplastic rubber dot pattern at 0.5-1.5 millimetres resists 60-160 newtons. The dot pattern is applied by screen printing at 0.12-0.48 USD per bed and it is the cheapest quality signal in the product.

Tolerance is the assembly issue. A cover cut to the foam's nominal dimension will not fit, because a foam block cut to plus or minus 2 millimetres on a 100 millimetre thickness is 2-4% out of square at the corner. The cover is cut 2-5% undersize to hold the foam in tension, and that undersize figure is the difference between a bed that looks tailored and one that sags.

A bolster of solid foam rather than loose fibre, a silicone dot base resisting 60-160 newtons, and a cover cut 2-5% undersize to hold the foam in tension, are the three construction decisions that make a bed look tailored rather than bagged.

Zipper, Seam and Wash-Cycle Durability

A bed that cannot be washed is replaced rather than cleaned, and the components that decide whether it can be washed are the zipper, the seam and the thread. All three are specified items and all three are frequently bought to price.

The zipper is a coil or a metal chain at 3-5 millimetres gauge, with a length of 400-900 millimetres depending on bed size. A coil zipper in polyester with a metal or a moulded slider costs 0.35-1.60 USD and survives 300-800 wash cycles; a metal chain costs 0.85-3.20 and survives 500-1,500 but it is heavier and it marks a washing machine drum.

Slider quality is the failure point rather than the chain. A slider without an auto-lock backs off under a load from inside the cover and the bed opens in use; an auto-lock slider costs 0.06-0.24 USD more and it is the difference between a zipper that stays closed and one that is reported as broken. The test is a pull of 40-90 newtons applied at the slider in the closed direction.

Seam type is the second decision. A plain lockstitch at 8-12 stitches per 25 millimetres is the standard and it is adequate on a straight seam; a flat-felled or an overlocked seam is needed where the cover is washed, because a raw edge frays in 3-15 cycles. An overlock costs 0.08-0.30 USD per bed and it is the specification for a washable cover.

Thread is the third and it is the one that fails invisibly. A polyester thread at Tex 30-50 survives 300-1,000 wash cycles; a cotton or a cotton-wrapped polyester thread degrades at 60-200 cycles and it shrinks, which opens the seam. The thread has to be polyester where a hot wash is specified, and the specification note is drawn rather than assumed.

Wash testing is the qualification. A cover is washed at 40-60 degrees for 10-30 cycles and measured for dimensional change of plus or minus 3%, for colour fastness to washing at grade 4 or above on a 1-5 grey scale, and for seam integrity after the final cycle. A cover at plus or minus 6% dimensional change will not go back on its foam after three washes.

Shrinkage interaction with the undersize cut is the detail that catches programmes out. A cover cut 3% undersize to hold the foam in tension, on a fabric that shrinks 3%, is 6% undersize after one wash and will not close. The fabric's residual shrinkage is specified at under 2% and it is measured before the cover pattern is graded.

An auto-lock slider tested at 40-90 newtons, an overlocked seam, a polyester thread at Tex 30-50, and a fabric with residual shrinkage under 2% before the pattern is graded, is the washable-cover specification.

Compression Packaging and Recovery After Unpacking

A bed is mostly air and it is freighted as air unless it is compressed. Compression is the single largest lever on landed cost in this category and it carries a recovery obligation that has to be engineered rather than assumed.

Vacuum compression removes 55-80% of the packed volume. A bed at 900 by 700 by 180 millimetres, or 0.113 cubic metres, packs at 0.023-0.051 cubic metres compressed, which is a factor of 2.2-4.9 on freight. On a long route at 3.20-11.60 USD per uncompressed unit, compression takes the freight to 0.70-5.30 and the saving is larger than the entire packaging cost.

Recovery is the obligation. A conventional or high-resilience polyurethane recovers 88-97% of its thickness within 4-24 hours of unpacking; a viscoelastic foam recovers 70-92% in the same window and may need 48-72 hours at 20 degrees or above. A bed that arrives and does not recover is a return, and the recovery figure belongs in the specification.

Compression dwell is the risk variable. A foam held compressed for 30 days recovers fully; one held for 90-180 days in a hot container recovers 82-94% and may take 48-96 hours. The control is a stated maximum compression dwell of 45-90 days in the shipping documentation, and a rotation rule for any stock held compressed beyond it.

Roll packing is the alternative to a flat vacuum pack. The bed is rolled around a core with the air expelled, giving a cylinder of 200-380 millimetres diameter and 400-900 millimetres length at 0.019-0.045 cubic metres. It recovers faster than a flat pack because the strain is distributed, at 92-99% in 2-12 hours, and it is the specification for a viscoelastic foam.

The bag is a specification item and not a consumable. A vacuum bag is a polyamide or polyethylene laminate at 80-160 microns with an oxygen barrier, heat sealed at 150-190 degrees and 1.5-3 seconds. A bag that loses its seal at 0.5-2% per day re-inflates in transit and the freight saving evaporates, which is why the seal is tested to a hold of 24 hours at 90% vacuum rather than to a visual check.

Carton arithmetic closes it. A compressed bed at 0.030 cubic metres and 1.6-4.2 kilograms packs 1,900-2,270 per 40-foot high-cube before pallet losses against 520-600 uncompressed. Weight never binds on a bed; volume always does, and every 10 millimetres of compressed thickness is worth 60-180 units per container.

Vacuum compression removes 55-80% of packed volume and cuts freight by a factor of 2.2-4.9, but it obliges a stated recovery figure of 88-97% in 4-24 hours, a maximum compression dwell of 45-90 days and a bag seal holding 90% vacuum for 24 hours.

Size Ladder, Cost Structure and Programme Economics

A bed range is a size ladder sharing one foam specification and one cover platform, and the economics of the range are set by how much of the bill of materials scales with size and how much does not.

Foam scales with volume and it is 40-62% of the material cost. A small bed at 600 by 450 by 100 millimetres uses 0.027 cubic metres of foam; an extra-large at 1,200 by 800 by 180 uses 0.173, which is 6.4 times the foam at 3.1 times the retail price. The largest size is therefore the least profitable per unit of freight and the most profitable per unit of material, and the two do not cancel.

Cover fabric scales with area and it is 22-38% of the material cost. The same ladder uses 0.42 square metres of fabric on the small and 1.62 on the extra-large. Sewing time rises from 3-7 minutes to 8-18. The zipper, the label and the packaging are the elements that do not scale, and they are 8-18% of the cost on the small size and 3-7% on the large.

The table below sets out the ladder on the parameters that decide the range structure.

The second cover line is the margin opportunity in the table. It uses the same pattern and the same fabric platform, it needs no foam and no testing, and it converts a one-time purchase into a repeat one at 3.20-12.60 USD against a 12-34 USD retail price.

The bolster line shows the honest cost of shape: 30-70% more foam and 25-55% more fabric for a 14-38 USD retail step, which is a thinner margin than the size ladder and which is why a bolster is offered on two sizes rather than four.

Compression is what makes the ladder work at all. Without it, the extra-large bed at 0.173 cubic metres ships 340-390 units per container and its freight is 9.20-28.40 USD, which is more than its material cost. With compression at 0.047 it ships 1,300-1,500 and the freight falls to 2.40-7.80, which is what makes the top of the ladder sellable.

Compression is what makes the size ladder viable: an extra-large bed ships 340-390 units per container uncompressed at 9.20-28.40 USD of freight, and 1,300-1,500 at 2.40-7.80 compressed, which is more than its entire material cost.