Those who innovate


Instant-Lab (EPFL): Breaking the second barrier

INNOVATION

Español Français
October 2020


Instant-Lab (EPFL): Breaking the second barrier

It is sometimes said that, as far as mechanical watch movements are concerned, everything, or nearly everything, has already been achieved. So is mechanical horology set in stone? Nothing could be farther from the truth. Mechanical watchmaking still has plenty of surprises in store for us, as demonstrated by Dr Ilan Vardi, senior scientist at the EPFL Patek Philippe Chair in Micromechanical and Horological Design, known more simply as: “Instant-Lab.”

I

nstant-Lab: an apt name for this laboratory of some twenty people which, under the supervision of Professor Simon Henein, specialises in the creation of new mechanisms with kinetic and technological innovations at the centimetre scale. The approach they apply is strictly scientific and is inspired by mechanical design in fields including horology, medical instruments and robotics.

Current Instant-Lab projects involve mechanical watchmaking and biomedical instruments, fields related by their technology and their industrial structure (recall the watch part suppliers working in Med-Tech, as perfectly reflected at the EPHJ).

Born in Paris, Ilan Vardi is a mathematician with a BSc from McGill University, Montreal, Canada and a PhD from MIT in the USA. He has published numerous research articles in mathematics, computer science, physics and the history of science. In the last 10 years he has specialised in horology, producing technical publications and patents on oscillators, escapements and complications. He has been Senior Scientist at EPFL Instant-Lab, Neuchâtel, since 2013.
Born in Paris, Ilan Vardi is a mathematician with a BSc from McGill University, Montreal, Canada and a PhD from MIT in the USA. He has published numerous research articles in mathematics, computer science, physics and the history of science. In the last 10 years he has specialised in horology, producing technical publications and patents on oscillators, escapements and complications. He has been Senior Scientist at EPFL Instant-Lab, Neuchâtel, since 2013.

But beyond its purely academic mission, the laboratory also aims to establish links with Swiss watchmaking culture and welcomes industrial collaboration with all Swiss watch companies, as stated by Instant-Lab. This is an important point, for, while Patek Philippe is the main sponsor of the Chair bearing its name, the multidisciplinary research carried out there serves the entire Swiss watchmaking community.

Passing through the second barrier

Dr Ilan Vardi is conducting advanced research in watch time bases. His goal, to put it simply, is to “pass through the second barrier” of the mechanical watch, in other words to take its accuracy to better than one second a day. A tough challenge indeed.

To achieve this, he explains, you have to replace the spiral- balance with an oscillator that has a significantly higher quality factor. But what exactly is the “quality factor,” introduced into watchmaking forty years ago by the British engineer Douglas Bateman?

EVOLUTION OF QUALITY FACTOR Q, FROM PIN PALLET BALANCE TO ATOMIC CLOCK
EVOLUTION OF QUALITY FACTOR Q, FROM PIN PALLET BALANCE TO ATOMIC CLOCK
The quality factor quantifies the energy lost by the oscillator at each oscillation. The principal cause is friction. In a spiral balance system, this quality factor is at best 300. It needs to be ten times better to break through the “second barrier.”

The quality factor quantifies the energy lost by the oscillator at each oscillation. The principal cause is friction. This quality factor is given by the number of oscillations made by an oscillator “in free mode” before its amplitude falls to just 4.3 percent of its original amplitude. In a spiral-balance system, this quality factor is at best 300. It needs to be ten times better to break through the “second barrier.”

By comparison, the quality factor of a standard quartz movement is around 100,000, or even one million for the most sophisticated quartz watches, such as the Calibre 0100 by Citizen, which is accurate to one second a year. So is the mechanical watch doomed to stay forever behind this second barrier?

Increasing the frequency?

The standard response is to raise the frequency, because experience has shown that a watch operating at 5Hz is often more accurate than a watch at 4Hz. But raising the frequency has its limits because of the friction inherent in the mechanical spiral-balance system, the pivot of which turns, with significant friction, on a jewel – a ruby.

Raising the frequency unquestionably improves accuracy, as proven by the 360Hz acoustic tuning fork, first produced in the Bulova Accutron of 1960, the quartz watch, invented in Neuchatel in 1967, running at 8,192Hz, which enabled it to surpass one-second-a-day accuracy, or the 8,388,608Hz of the Citizen watch – 1s/year – not to mention the caesium atomic clock which, with a frequency of 9,192,631,770Hz (which, in fact, defines the second!), achieves an accuracy of 1s/100,000 years. Yet frequency is not the sole answer.

The importance of Q

Let us return to the quality factor, or Q for short. Remember that Q essentially represents the number of free oscillations before the oscillator stops. For wristwatches, the average is about 200. But a marine chronometer spiralbalance, which beats at the same standard frequency of 4Hz, has a heavier balance and so a Q factor of the order of 1,000. It is ten times more accurate, attaining 1s/week. Note that, unlike wristwatches, marine chronometer movements remain horizontal as they are mounted on gimbals.

Another example corroborating the importance of Q is the pendulum clock. The best of them, the Shortt clock dating back to 1921, achieves the same 1s/year accuracy as the Citizen Calibre 0100. And yet its frequency is only 0.5Hz, as compared to the 8 million Hz of the Citizen watch. The key is that its Q factor is very high, surpassing 100,000.

The Shortt clock, of which around one hundred copies were produced between 1922 and 1956, keeps time by means of two oscillators, a master pendulum encapsulated in a vacuum and a slave pendulum in a separate clock, which are synchronised by means of an electromagnetic mechanism. The slave pendulum is attached to the clock's timekeeping mechanisms, which leaves the master pendulum virtually free from external disturbances. This copy was sold by the specialist auction house Gardiner Houlgate for £25,000.
The Shortt clock, of which around one hundred copies were produced between 1922 and 1956, keeps time by means of two oscillators, a master pendulum encapsulated in a vacuum and a slave pendulum in a separate clock, which are synchronised by means of an electromagnetic mechanism. The slave pendulum is attached to the clock’s timekeeping mechanisms, which leaves the master pendulum virtually free from external disturbances. This copy was sold by the specialist auction house Gardiner Houlgate for £25,000.

The significance of Q was essentially stated by John Harrison, inventor of the first marine chronometer 250 years ago: “The pendulum must have dominion over the clockwork.” What he meant was that any disturbance to the oscillator leads to timekeeping errors, so disturbances such as impulses from an escapement must be minimised.

“The higher the quality factor, the more we can reduce the impulses required to maintain the oscillator. For example, mechanical watches provide an impulse to the balance wheel at every beat, whereas marine chronometers do so only on alternate beats and the Shortt clock only once every thirty beats!” Ilan Vardi explains.

The IsoSpring project

At Instant-Lab, the convergence of mathematician Ilan Vardi and engineer Simon Henein led to the IsoSpring project. The object of this programme is to “solve the problem of finding the best escapement by completely eliminating it.” The project begins by going back to Isaac Newton, who in 1687 imagined a hypothetical planetary system in which all planets rotated in elliptical orbits, but where years were the same for every planet.

Instant-Lab’s achievement was to realise this mechanically, leading to oscillators having isochronism, the key to accurate timekeeping, but replacing the stop-and-go oscillations of previous oscillators with continuous unidirectional orbits. “We went back to the 17th century when clock and watch oscillators were first invented and found one which we developed with modern engineering technology,” adds Ilan Vardi.

The first IsoSpring prototype, demonstrated publicly in February 2014
The first IsoSpring prototype, demonstrated publicly in February 2014

The key to Q

As Dr Ilan Vardi explained in a 2014 article: “The quality factor of the oscillator is scientifically recognised as being the best indicator of timekeeper accuracy. This is why the current race for high frequency should rather be a race for high quality.”

So what are the keys to raising the quality factor? “You have to minimise the friction between solids: removing pivots raises Q by a factor of 10. You then minimise internal friction: replacing metal with silicon raises Q by another factor of 10. Lastly, you minimise air friction: sealing the oscillator in a vacuum raises Q by yet another factor of 10,” explains Dr Vardi.

Flexure bearings

Flexure bearings, also known as compliant mechanisms, use springs to provide rotational motion, instead of the traditional pivot. Since solid friction is eliminated, this significantly improves the Q factor. These were used in IsoSpring oscillators, so in combination with a monobloc construction in silicon, the Q factor was increased to 3,000.

The elliptical unidirectional orbits of the oscillators meant that the traditional escapement mechanism could be replaced by a simple crank. The result was a revolutionary oscillator with promising chronometric performance. Several prototypes were designed during these five years of development. One problem remains: sensitivity to gravity.

Spherical IsoSpring
Spherical IsoSpring
In December 2016, the first IsoSpring clock was unveiled at Neuchâtel City Hall. Its revolutionary oscillator is the time base for a mechanical clock powered by a driving weight. Its case is inspired by the traditional Neuchâteloise clock. Highly symbolic. “Because invention in watchmaking has to be revolutionary and traditional,” explains Ilan Vardi, savouring his words. “That sums up IsoSpring. It's totally innovative, but has a rich scientific and technical backstory.”
In December 2016, the first IsoSpring clock was unveiled at Neuchâtel City Hall. Its revolutionary oscillator is the time base for a mechanical clock powered by a driving weight. Its case is inspired by the traditional Neuchâteloise clock. Highly symbolic. “Because invention in watchmaking has to be revolutionary and traditional,” explains Ilan Vardi, savouring his words. “That sums up IsoSpring. It’s totally innovative, but has a rich scientific and technical backstory.”

Escaping gravity in a wristwatch

The challenge now is to miniaturise the system to the scale of a wristwatch. But as the first IsoSpring oscillators were sensitive to gravity, it was impossible to implement them in a wristwatch. To obtain an oscillator insensitive to gravity, Instant-Lab developed a new, patented mechanism that they named the Wattwins oscillator, a reference to its historical inspiration – ‘Watt’s linkage’, also known as the parallel linkage, invented by James Watt in 1784 in order to obtain straight line movement for steam engine pistons.

Similarly, the Wattwins oscillator consists of four flexure bearing oscillators whose rotations are converted into linear motion according to Watt’s design. The Wattwins oscillator is thus insensitive to linear acceleration, which is crucial in a wristwatch, as well as to angular acceleration – which is not the case for mechanical watches. Prototypes were developed at the watch scale to demonstrate the insensitivity to gravity of these IsoSpring oscillators.

Wattwins oscillator in silicon at the scale of a wristwatch
Wattwins oscillator in silicon at the scale of a wristwatch

The stage is now set to develop a mechanical wristwatch capable of “breaking the second barrier.” But, as Ilan Vardi points out: “Our laboratory is not involved in the industrial phase. We leave that part to the manufactures, who have far more resources and know-how in that field. We also leave it to the companies to sort out the details of perfecting our prototypes.”

Is this the royal road to restoring the antiquated mechanical watch to its former glory? The answer will depend on how successful this line of research proves to be.