automatic sliding door with lock

Ask ten suppliers to quote an automatic sliding door with a lock and you will get ten quotations that are not quoting the same thing. Not because nine of them are being dishonest, but because the sentence contains a word that does nine different jobs in this industry, and a second word that describes a state rather than a part. Nobody in the chain is lying. Everybody in the chain is answering a different question.

On a manual sliding door, a lock is a component. You buy it, it goes in the stile, and it does or does not have a bolt in the keep. On a powered door, a lock is a condition the door has been put into. It exists in one mode out of six, under one of two power states, and - if the door happens to be on an escape route - only after a named person has decided something about the building that has nothing to do with the door.

The clearest evidence for this is a single sentence in the user manual of one European sliding door operator. Describing the mode that closes the door at the end of the day, it says: in an escape route the OFF mode may only be set after it is certain that all people have left the building. That is not a sentence about hardware. It is a sentence about a headcount. The mode that locks your door is the only door mode whose lawfulness depends on whether the building is empty, and no lock manufacturer can sell you the answer to it.

What follows is written for the person who has to specify this once and live with it for a decade. We make the aluminium leaf. We do not make the operator, we do not make the access control lock, and we will say plainly which sentences belong to which party.

The word "lock" is doing nine jobs in this sentence

Before you compare quotations, separate the nine things that get called a lock. Four of them are not locks at all. Two of them are locks that are designed to lose. One of them is a lock on the decision rather than on the leaf. The table below is the index for the rest of the article.

Notice that the two rows most buyers are actually asking for - rows one and two - are not locks at all. This is why the phrase sliding door with lock produces quotations that differ by more than the price of a lock: half of them are quoting a mode, and a mode is a software setting in somebody else's controller.

Six modes, and the lock only exists in one of them

Every powered sliding door has a mode selector. The names differ between manufacturers, which is itself worth knowing, because a buyer who has read one manual will assume the word means what it meant there. We have put two manufacturers' published mode tables side by side below and deliberately not reconciled them. Where they disagree, that disagreement is the finding.

Two things in that table deserve to be said out loud, because nobody puts them in a quotation.

First: the mode in which you are permitted to lock the door is the mode in which the door stops looking. "All activation and safety devices are isolated once the door is fully closed" is not a footnote. It means the presence sensors that protect people in the doorway during the day are not protecting anything at night. If your night-time risk model assumed the door still had its sensors, it does not.

Second: one manufacturer's display does not show one locked state. It shows two characters - one while the door is travelling to the locked position, one once it has arrived and the lock is confirmed. A two-character display for two states that buyers collapse into a single word is a fair measure of how much is being assumed away.

The same power cut, four different doors

If a lock were a property of the door, then a power failure would produce one outcome. It does not. At least one European sliding door operator installation manual publishes a whole chapter on mains failure, and the outcome depends on which mode the door was in when the lights went out, whether an emergency battery is fitted, and whether the operator has mechanical emergency opening.

There is no honest way to describe a lock as a property of a door that behaves in four different ways on the same event. What you are buying is a set of configured behaviours, and the configuration is done by whoever commissions the door.

Fail safe and fail secure are about people versus assets, not about quality

The single most misread pair of words in access control. "Fail safe" sounds like the stronger specification. It is the weaker one for your stockroom and the only lawful one for your escape route, because the word safe refers to the person trying to leave, not to the thing you are protecting.

We are not going to tell you which to choose, and we would distrust any door supplier who does. That is a decision for your fire engineer and your authority having jurisdiction. What we will tell you is the question to ask: is this door on a required escape route, in writing, signed by somebody who owns that decision? Until that is answered, every lock quotation is a guess.

One word, three terminals, and the third one is the one that matters

Look at the input and output list of a published sliding door controller and the word lock appears more than once, on separate terminals, doing separate jobs. This is the clearest published proof that a lock is a process and not a part.

There is a matching detail in the service manual of a different class of sliding door operator - the sort used where the consequence of a failed lock is not commercial. It describes adjusting the door position switches so that they change state just as or slightly after the leaf stops moving, but before the carriage finishes its final travel, "allowing the device motor to coast into the full locked position". And then: the device should provide a secure indication only when the door is closed and the lock bar is completely down.

That is the whole problem in one sentence. The mechanical lock completes after the motor has stopped driving. The motor coasts into the lock. If the leaf is a millimetre short - a swollen gasket, a cold morning, a grain of sand in the track - the bolt has nothing to go into, and the system reports "closed" rather than "locked". Your log will say the building was secured. It was not.

Two locks on one leaf is not redundancy. It is a fault

Here is the collision that nobody prices in, and it comes straight out of the operator manuals rather than out of anything we invented. One manufacturer prints it as an IMPORTANT NOTICE in block capitals: if manual locks are fitted to the door leaf, ensure they are UNLOCKED before switching the drive into any operational mode, and switch the drive to night mode before locking the door. We found the same notice in two of that manufacturer's manuals, for two different drive families.

Now hold that against our own catalogue. One of the sliding door series we publish - the SL126/182 variant with the seventy five millimetre sash - lists a multi-point locking design with stainless steel locks in precision-cast reinforced casings. That is a real lock, in a real stile, and we are glad it is there. But if that leaf is motorised, that lock becomes precisely the device the operator manual tells you to release first.

The last row of that table is the one that matters most from us. If you are buying a leaf from a system house and an operator from an automation house and a lock from an access control house, then the sentence "who releases what, and in what order, and who proves it" belongs to none of the three. It belongs to the commissioning engineer, and it has to be written down before the leaf is cut.

Where the lock physically goes, and why that is a factory decision

There is a granted United States patent for an electromagnetic lock assembly for sliding double doors. We are not interested in the invention. We are interested in its background section, because a patent applicant has to describe the problem honestly in order to justify a solution, and what it describes is the most useful published paragraph we have found on this subject.

• "Closing the sliding double door does not lock the doors. Although the sliding double doors are closed, they can be forced open."
• "All electromagnetic locks are designed for mounting to a stable, non-moveable, door frame."
• "Electromagnetic locks are seldom installed on sliding double automatic doors because both doors move away from each other in operation, and the sliding doors have an aluminium access panel above the two doors that is fabricated from very thin metal, which provides little support for an electromagnetic lock system."
• "Installation is extremely difficult because the electronics in the door can be damaged during installation. Therefore, careful drilling of the door is required."
• "To date, in order to lock this type of sliding double doors, an individual must physically lock the doors with a key."

Read that third bullet again. The reason a magnetic lock is hard to fit to a powered sliding door is not the lock. It is the aluminium above the door. The transom is a cover, and a thin one, and a lock that is asked to resist a pry bar cannot be anchored to a cover.

That is a fabrication decision, and it is the one thing in this whole article that a door system supplier genuinely owns. Reinforcement for a lock body, a machined pocket instead of a drilled hole, a keep plate that is part of the section rather than screwed on afterwards, and a cable raceway that does not require somebody to drill through a transom full of wiring on a ladder. All of those have to be in the extrusion and the machining file. None of them can be added sensibly after the leaf is glazed.

The building side of it is just as unforgiving. One operator installation manual states that the transom must be fastened to a flat surface solid enough to bear the weight of the leaves, that the fastening points must be distributed alternately along the reference lines every three hundred millimetres with M6 or equivalent fixings, and that the transom is not self-bearing. A lock that holds a leaf still is asking the structure behind the transom to take a load in a direction the operator never asked for.

Reading our own published numbers for this question

We publish a specification table for each sliding door series. Almost none of it is about locking, and the parts that look relevant are usually measuring something else. Here is the honest reading, including the two places where our own table is easy to misread.

If you take one row away, take the sash width row. The number that gets quoted back to you in a brochure is the frame width. The number that decides whether the lock you have specified will physically go into the door is the sash width, and on our narrowest published sliding door sash that is forty four and a half millimetres. A security sliding door is not a product category; it is a set of decisions about which millimetres go to the sight line and which go to the hardware.

The escape route question is not written on the door

Everything above assumed one door. In practice there are two doors with identical hardware and opposite lawful behaviour, and the difference is a decision made on a fire strategy drawing by somebody who never sees the quotation.

That last point about key or code deserves emphasis, because it is the least known and the most useful. A European reading of EN 16005 does not ban locking a powered sliding door on an escape route. It restricts how you get into the locked state. A time clock that locks the entrance every night at eight is a very normal thing to specify and, on that reading, the wrong thing to specify. What is wanted is a deliberate act by a person who is accountable for the building being empty.

The second lock: the switch that decides whether the leaf is locked

Once you accept that locking is a mode and not a part, a second lock appears in the specification: the lock on the mode selector. One manufacturer's published brochure states that automatic sliding doors in escape and rescue routes must be secured against operation by unauthorised persons, and offers a lockable mechanical switch, a key switch, or a code. Another's user manual adds the small, hard instruction that the key must always be removed from the mode selector on emergency escape doors after changing the settings.

Nobody quotes for this. It is a few pounds of hardware and it is the difference between a locked entrance and a locked entrance that a passer-by can unlock by pressing a button. If it is not in the schedule, it will not be on the wall.

Break-out is a lock that is designed to lose, and losing costs a visit

On an escape route, one of two things must be true: either the leaves can be pushed out of their track, or the operator is of a type that opens them on power failure without being pushed. Both are locks designed to lose, and the cost of losing is not nothing.

A published guide to resetting a break-out door makes the point that a break-out reset is not a power reset. Cycling the supply does not restore a door that a break-out switch has told the controller is no longer in its normal running configuration. The leaves have to be re-seated, the latches re-engaged, the break-out switch cleared, and the sequence followed - and the sequence is manufacturer-specific, so a generic button combination can create a second fault. The same guide's list of when to stop and call an engineer is short and worth reading before the day you need it.

The practical instruction for a facilities team is simple and almost never written down: put the reset procedure and the responsible person in the handover file, not in the service contract. A door that has been broken out and left inhibited is a door that is neither automatic nor locked, and it will sit like that until somebody notices.

Reduced opening width on an escape route has a floor

The winter setting is the most harmless-looking feature in the whole controller and it has a number attached. One operator manual states that the reduced opening width may not be less than thirty percent of the opening width, and that in emergency exits the reduced opening width must be at least as large as the required escape route width. Another states that with an emergency push-button a partially opening door opens fully.

So the sequence is: somebody sets a winter opening to save heating, the leaf now opens to a fraction of its travel, and the escape capacity of the entrance has changed without anybody touching the door. If the entrance is a required exit, that setting is a fire strategy input, not a facilities preference. Write it into the commissioning record and lock the mode selector.

What we can put on your drawing, and what we cannot

We would rather say that than let you build a schedule on a sentence we cannot support. Our published catalogue is four manual sliding door series. If your project needs a powered entrance, the honest shape of the supply chain is: a leaf from us, an operator from an automation house, a lock from an access control house, and one named person who owns the sequence.

Twelve lines to send with the enquiry

Most of the failures described in this article happen because a question was never asked, not because an answer was wrong. Send these twelve lines to every supplier in the chain - leaf, operator and lock - and compare what comes back. The last column is the one that saves you.

What we publish, and what you still have to ask for

Six questions people actually ask

Is a locked automatic sliding door on an escape route ever allowed

Yes, in the European reading we have cited, and the restriction is on the method rather than on the fact of locking: it may be locked only by key or code, meaning only by an authorised person. What is not acceptable is a lock that would hold against someone trying to leave, and what is not acceptable in most jurisdictions is getting into the locked state by a clock rather than by a decision. Your authority having jurisdiction is the only body that can answer for your building.

Can I just add a magnetic lock to the sliding door I already have

The published background of a granted patent on exactly this problem says these locks are seldom installed on powered double sliding doors, for two reasons that will still be true on your door: both leaves move, and the aluminium panel above them is thin and gives the lock little support. There is also the drilling problem - the same source warns that the electronics can be damaged during installation. If you want to do this, the reinforcement should have been designed into the section. It is a much cheaper conversation before the leaf is made than after.

Why does my system report the door as secured when it obviously is not

Because it is reporting a command rather than a state. A published controller has separate terminals for the lock output, the lock position feedback input, and the emergency unlock input. If the feedback is not wired and monitored, the system is showing you what it asked for. The related failure is mechanical: on one service manual the lock completes after the drive has stopped and the motor coasts into the locked position, so a leaf that stops a millimetre short closes but never locks.

What is the difference between exit-only and locked

Exit-only is a direction. The outside activator is disabled and the inside one still works, so the building empties and does not fill. It is not a bolt, and one manufacturer's manual makes the distinction explicit by saying the door is normally locked in that mode if a locking device has been fitted - and unlocked if it is an escape route. The same switch position, two opposite behaviours, decided by a drawing you have probably never seen.

Do the safety sensors still protect people at night

No, and this is stated plainly in at least one operator manual: all activation and safety devices are isolated once the door is fully closed in night mode. That is the point of the mode - the door is shut and not cycling. But it means any assumption that the doorway is still monitored outside opening hours is wrong, and it is another reason the decision to enter that mode should belong to a person rather than to a timer.

The door was pushed open in an emergency and now will not run. Why did a manual release break the automation

Because a break-out is not a power event. A published reset guide makes the point that cycling the supply does not clear it: a break-out switch has told the controller the door is no longer in its normal running configuration, and the leaves, latches and operator all have to be returned before the controller will accept a reset. The sequence is manufacturer specific. Put it in the handover file with a named person against it.

The short version

• Nine different things get called a lock on a powered sliding door. Four of them are not locks, two are locks designed to lose, and one is a lock on the mode selector rather than on the leaf.
• A lock is a state, not a part. It exists in one of six modes, and the same power failure produces four different outcomes depending on which mode the door was in.
• The mode in which you are allowed to lock the door is the mode in which the safety devices are switched off.
• Fail safe means safe for the person leaving, not strong. Fail secure means locked when the power dies. Neither is better; they answer different questions.
• One controller gives you three separate terminals for the word lock: the command, the proof, and the release. The one that gets skipped is the proof, and the one that saves you is the release.
• A manual lock in the leaf and an electric lock in the operator are not redundancy. Two operator manuals say to unlock the manual one before switching the drive on - and one of our own sliding door series publishes a multi-point lock that would be exactly that manual lock.
• Where the lock goes is a fabrication decision. A magnetic lock's design assumes a frame that does not move, and the thin aluminium panel above a sliding door is not a fixing surface.
• In our catalogue, the number that decides whether a lock fits is the sash width, from forty four and a half to ninety six millimetres. The number that gets quoted is the frame width. They are not the same number.
• Four hundred kilograms and two hundred and twenty kilograms are two different measurements on two different series. Do not put them in one column.
• Whether the door is on an escape route changes the lawful behaviour of identical hardware, and one European reading restricts locking there to key or code - a person, not a schedule.
• A break-out is a lock designed to lose, and losing it costs a service visit, not just a push. Put the reset procedure in the handover file.
• We make the leaf. We do not make the operator or the lock, we will not sign the escape route approval, and we would rather tell you that than let you build a schedule on it.