Why Barn Door Tracks Sag at the End

By Evan Christensen, co-owner of The Barn Door Hardware Store and a former hardware structural simulation engineer. The deflection figures on this page are his own calculations.

If a barn door dips when it's slid fully open, or a pair of biparting doors meets at the top and leaves a wedge of daylight at the bottom, the doors are usually not the problem. The track is.

A barn door track is a structural beam carrying a rolling load, and beams don't fail in the middle of a clean span. They fail at the two places the load path gets interrupted: the ends and the splices. Barn door hardware is usually sold on finish and price, and neither tells you anything about what happens at those two points.

This page works through both, with geometry you can check yourself on any track — including ours. Goldberg Brothers, the line we carry, engineers both points. Our own Quick Ship line does not, and we will be specific about that below. The mechanism is checkable either way, so you shouldn't have to take our word for it.

The ends: what happens when the door parks

Slide a door to the end of its track and the entire weight of that door now sits at the extreme end of the beam. What holds it there depends on where the last mounting hole is.

On many tracks, the last hole sits several inches inboard of the end. Everything past that point is cantilevered — unsupported steel holding a live load. On Goldberg track, there is a lag bolt located at the end of the track. The load passes directly into the header board instead of out onto an overhang.

Why a small amount of deflection shows up as a large gap

Here is the part most people miss. A door hangs from two rollers, and the distance between those rollers is far smaller than the height of the door. That geometry is a lever. If one roller drops while the other stays put, the door doesn't just sink — it rotates, and the rotation gets multiplied down the height of the door.

The math is simple: a door amplifies roller drop by its height divided by its roller spacing. Take a common 36″ × 84″ door with its hangers set 4″ in from each edge, which puts 28″ between the rollers. That is 84 divided by 28, or 3×. Put two doors nose to nose in a biparting setup and both of them tilt toward each other, so the visible effect doubles to 6×.

A track that moves 1/16″ opens a 3/8″ gap between a pair of biparting doors. The doors touch at the top and gap at the bottom — the single most common complaint we hear about biparting installations, and it is almost never the doors' fault.

What the numbers actually say

Goldberg track is bolted at the end, so a cantilevered end is not a condition it has. To show what that end bolt is preventing, the figures below model a track without one.

The section modeled is a 1/4″ × 1-1/2″ steel flat bar. That is the section our own Quick Ship line uses. We picked a thicker bar on purpose rather than a thin one, because it makes the result harder to argue with — and because it is the one we can speak to with certainty.

The numbers assume:

• Door: 36″ wide × 84″ tall, hangers set 4″ in from each edge, giving 28″ between rollers
• Load: half the door's weight carried by the roller nearest the end when the door is parked there
• Track fastened into framing at 16″ on center
• Steel, E = 29,000,000 psi

Each figure is a range. The low end treats the track as rigidly fixed at the last lag bolt; the high end lets that bolt rotate freely. Real framing sits somewhere between the two.

Door weight	Overhang past the last bolt	Track drop at the roller	Gap at the bottom of a biparting pair
100 lb	3″	0.000″ to 0.001″	under 1/64″ — invisible
100 lb	6″	0.002″ to 0.007″	about 1/64″ to 1/32″ — borderline
100 lb	9″	0.006″ to 0.017″	about 1/32″ to 3/32″ — visible
100 lb	12″	0.014″ to 0.033″	about 1/16″ to 3/16″ — obvious
150 lb	12″	0.021″ to 0.049″	about 1/8″ to 5/16″ — obvious

Three things fall out of this that are worth saying plainly.

Short overhangs are fine. If the last hole is within about 3″ of the end, the steel barely notices. We are not going to tell you a 3″ overhang will ruin your install, because it won't.

Deflection scales with the cube of the overhang. Double the unsupported length and you get eight times the movement. That is why this goes from a non-issue to an obvious defect over a span of just a few inches, and it is why the location of one hole is worth caring about.

Thicker steel does not save you. That 1/4″ × 1-1/2″ bar has a stiffness of 0.0703 in⁴. Goldberg's standard-duty track is thinner — 3/16″ × 1-1/2″, or 0.0527 in⁴ — so the bar modeled above is about 33% stiffer in bending. It still performs worse at the end of a run, because what matters is how far the load sits past the last fastener, not how heavy the steel is. This is the whole argument in one line: support beats section.

The deflection above is elastic — the steel springs back when the door moves away. That is not reassuring, though, because the door is only at the end when it's closed, which is exactly when someone is standing there looking at it.

There is a second, slower mechanism we can describe but won't put a number on: a cantilevered end applies a prying moment to the last fastener every time the door parks. At a 12″ overhang with a 50 lb roller load, that's roughly 600 in-lb of leverage working on one lag bolt, cycled several times a day for years. Fasteners in wood don't like repeated prying. We haven't run a fatigue study on it, so we'll leave it as a mechanism rather than a claim.

Run the numbers on your own opening

None of this is proprietary. It's standard beam theory, and you can check it on any track you're looking at, ours included. Three steps.

Step 1 — find the stiffness of the track section.
Formula: I = b × h³ ÷ 12
b = thickness of the bar, in inches
h = height of the bar, in inches
Result: I, in inches to the fourth power (in⁴), the standard unit for a section's resistance to bending
Example: a 1/4″ × 1-1/2″ bar → 0.25 × 1.5³ ÷ 12 = 0.0703 in⁴

Step 2 — find how far the track end drops.
Formula: d = P × a³ ÷ (3 × E × I)
P = load on the roller nearest the end, in pounds (roughly half the door's weight)
a = overhang, measured from the center of the last lag bolt to the end of the track, in inches
E = 29,000,000 psi for steel
Example: 50 lb on a 12″ overhang → 50 × 1,728 ÷ (3 × 29,000,000 × 0.0703) = 0.014″

Step 3 — convert that into what you'll actually see.
Formula: gap = 2 × d × H ÷ s
H = door height, in inches
s = roller spacing, in inches
Drop the 2 for a single door; keep it for a biparting pair, where both doors tilt
Example: 2 × 0.014 × 84 ÷ 28 = 0.085″, or about 3/32″

Step 2 assumes the track is rigidly fixed at the last bolt, which is the optimistic end of the range. If that bolt can rotate at all — and in real framing it can — the answer grows. Treat what you calculate as a floor, not a ceiling.

Why the overhang varies at all

Everything in that table turns on one number: how far the last bolt sits from the end of the track. So it is worth asking what decides that number — and on our Quick Ship track, nothing does.

Our Quick Ship track is punched on 16″ centers so the holes land on studs. That is sensible for mounting, and it is why a track goes up quickly and lands in solid framing. But it means the hole pattern is laid out from the wall, not from the track. The last hole falls wherever the sequence happens to run out, and the steel past it is whatever is left over. Depending on the length you order, that remainder can be almost nothing or almost a full 16″ bay.

That is not a defect. It is what happens when a mounting pattern is designed around framing and the ends of the track are left to fall where they fall. We can only speak for our own track here; if you are comparing others, the check at the end of this page tells you what to measure on them. The consequence is that the outcome is unpredictable: two installs of the same product, in two different openings, can land in different rows of the table above.

Here is the span, for that 1/4″ × 1-1/2″ section carrying a biparting pair:

• 1″ of overhang: nothing you could measure
• 8″ of overhang: roughly 1/32″ to 1/16″ at the bottom of the doors
• 16″ of overhang, 100 lb doors: roughly 3/16″ to 3/8″
• 16″ of overhang, 150 lb doors: roughly 5/16″ to 5/8″

Goldberg Brothers places a fastener at the end of the track regardless of its length. That does not make the steel stiffer — as the numbers above show, the section has remarkably little to do with it. What it does is remove the variable. The overhang is not short or long depending on which opening you happen to have. There isn't one.

Which tier your opening actually needs

We sell both, and we would rather you bought the right one than the expensive one.

Our Quick Ship line uses the 1/4″ × 1-1/2″ section, punched on the stud grid, with no bolt at the end. It ships from stock and it costs less. On a short run where the open door parks close to the last hole, it behaves no differently from anything else described on this page. The table says so, and we are not going to pretend otherwise to sell you something dearer.

The end bolt earns its price in four situations:

• Biparting pairs. Both doors park at the center, both tilt toward each other, and the geometry multiplies whatever the track does by six.
• Long runs. More track means more opportunity for the hole pattern to leave a long remainder at the end.
• Heavy doors. Deflection scales directly with weight, so a 150 lb door does half again what a 100 lb door does.
• Any opening where the fully open door parks past the last hole.

If none of those describe your opening, Quick Ship is a sound buy and it ships from stock. If any of them do, the Goldberg Brothers line is the one we would put in our own house, and the reason is the bolt rather than the badge.

The splices: a joint that keys together

Tracks longer than 7′6″ ship in sections, and how those sections rejoin determines whether you have one continuous beam or two beams in a trench coat.

A flat plate bolted across a square butt joint transfers load only through its fasteners. Mechanically that behaves closer to a hinge than to continuous steel — the two sections can shift slightly relative to each other, and any vertical shift leaves a step in the running edge that the wheel has to climb every pass.

Goldberg's split track doesn't meet on a straight line. Each section is cut with a pair of opposed wedges that hook into the matching recesses in its neighbor, so the seam zigzags back and forth through the middle of the bar instead of crossing it square. Those angled faces bear directly against one another: for either section to drop relative to the other, it would have to ride up the wedge. A bolted plate behind the joint clamps the two halves together, but the alignment is held by the steel itself, not by the bolts.

The seam still crosses the running edge squarely at the top and bottom of the bar — the interlock sits in the middle of the height, not at the edges. What the interlock does is keep those two edges from stepping out of line with each other, which is what turns a joint into a bump.

Worth knowing how this is applied, because it differs by configuration:

• Single doors on tracks over 7′6″ use the interlocking split track.
• Biparting setups over 7′6″ use a traditional connecting plate — and both tracks carry the end bolt. This is the configuration where the end bolt matters most, because the doors meet where the two tracks meet, so both doors are parked at the joint at the same time.
• Tracks at or under 7′6″ are single-piece. No joint at all.

One note of fairness: Goldberg introduced split track to reduce freight cost on long runs, and it does that. The engineering in the joint profile is a consequence of doing it properly rather than the original motivation. We think it's the more interesting half of the story.

Made to order, which is slower on purpose

Nearly all of our Goldberg Brothers kits do not exist until you order them.

That is the honest explanation for the lead time on those kits, and it is also what makes inspection meaningful. There are two exceptions, both deliberate: a small number of Goldberg kits we hold in stock so they can leave the next business day, and the Quick Ship line, stocked for the same reason. On a made-to-order kit, parts are cut and finished for your order, checked individually, and the assembled order is inspected against the packing list before the box is closed. The stocked kits are drawn from inventory we assembled and checked the same way, not from pre-boxed kits sealed months ago in a container.

Every hardware company on the internet claims it inspects things. The difference worth judging us on is not the claim — it is whether we're willing to be slower because of it. We are.

How to check any track before you buy it

You don't have to take our word for any of the above. Pull up the product photos for any track you're considering, ours included:

1. Find the last mounting hole. Measure from the center of that hole to the end of the track. Under about 3″, you're fine. Approaching a foot, take the table above seriously.
2. Look at how the joint is cut on any multi-section track. Square butt ends with a plate over them behave differently than an interlocking profile.
3. Ask what the track section is. A supplier that can't tell you the thickness and height of the steel hasn't thought about the beam at all.
4. Ask where the door parks. If a fully open door sits past the last fastener, you're loading an overhang every single day.

These are the four questions we'd ask, and they're the reason we carry the line we carry.

Common questions

Does every Goldberg track have a joint in it?
No. Tracks at or under 7′6″ ship in a single piece with no joint at all. The split track and connecting plate questions only apply to longer runs.

My biparting doors touch at the top and gap at the bottom. Is that a door problem?
Usually not. That specific pattern — contact at the top, wedge-shaped gap at the bottom — is the signature of the doors rotating rather than the doors being cut wrong. Check whether the track is supported at the ends where the doors meet before you plane anything.

Is the interlocking joint as strong as a single piece of track?
A joint is never identical to continuous steel. The interlocking profile's purpose is to hold the two sections in alignment through the geometry of the cut rather than through the fasteners alone. If your opening allows a single-piece track, that remains the simplest structure.

Do I need heavy duty track?
Goldberg's heavy-duty section (3/16″ × 2″) is about 2.4× as stiff in bending as the standard-duty section (3/16″ × 1-1/2″), because stiffness rises with the cube of the bar's height. That matters most on long spans, heavy doors, and biparting pairs where deflection is amplified. For a typical single door under 100 lb on a properly supported track, standard duty is not a compromise.

Why does my order take longer than the marketplace listings?
On a made-to-order kit, because the parts are made and finished for your order rather than pulled from pre-boxed stock — the lead time and the inspection are the same fact described two ways. If you need hardware sooner, the stocked Goldberg kits and the Quick Ship line both ship without that wait.

---

Deflection figures on this page were calculated in-house using standard beam theory (E = 29,000,000 psi) for a 1/4″ × 1-1/2″ steel flat bar with the overhang modeled between fixed and freely rotating end conditions, carrying a 36″ × 84″ door on 28″ roller spacing with mounting into framing at 16″ on center. They describe elastic deflection under a static parked load and are intended to show how the geometry behaves, not to serve as an engineering specification for any particular installation. Framing condition, fastener quality, and header construction all affect real-world results.