Scaling Laws and Multiscale

Manufacturing
 Scaling Laws
 In this era of “think small,” one would intuitively simply scale down
the size of all components to a device to make it small. Unfortunately,
the reality does not work out that way.

 It is true that nothing is there to stop one from down sizing the device
components to make the device small. There are, however, serious
physical consequences of scaling down many physical quantities.

 This topic will present, with a selected cases, the scaling laws that will
make engineers aware of both positive and negative physical
consequences of scaling down machines and devices.

24-Jun-20 IIT Jodhpur 2

 Content
 Scaling in Geometry

 Scaling in Rigid-Body Dynamics

 Scaling in Electrostatic Forces

 Scaling in Electromagnetic Forces

 Scaling in Electricity

 Scaling in Fluid Mechanics

24-Jun-20 IIT Jodhpur 3

 Why Scaling Laws ?
 Miniaturizing machines and physical systems is an ongoing effort in
human civilization.
 This effort has been intensified in recent years as market demands for:

 Intelligent, Robust, Multi-functional and Low cost consumer products

 The only solution to produce these consumer products is to package

many components into the product making it necessary to miniaturize
each individual components.

 Miniaturization of physical systems is a lot more than just scaling down

device components in sizes.

 Some physical systems either cannot be scaled down favorably, or

cannot be scaled down at all!

24-Jun-20 IIT Jodhpur 4

Types of Scaling Laws

1. Scaling in Geometry:

Scaling of physical size of objects

2. Scaling of Phenomenological Behavior

Scaling of both size and material characterizations

24-Jun-20 IIT Jodhpur 5

 Scaling In Geometry
 Volume (V) and surface (S) are two physical parameters
that are frequently involved in machine design.

 Volume leads to the mass and weight of device

components.

 Volume relates to both mechanical and thermal inertia.

 Thermal inertia is a measure on how fast we can heat or cool a

solid. It is an important parameter in the design of a thermally
actuated device in various microfluidics domain.

 Surface is related to pressure and the buoyant forces in fluid

mechanics. For instance, surface pumping by using piezoelectric
means is a practical way for driving fluids flow in capillary
conduits.
24-Jun-20 IIT Jodhpur 6
 When the physical quantity is to be miniaturized, the
design engineer must weigh the magnitudes of the
possible consequences from the reduction on both the
volume and surface of the particular device.

24-Jun-20 IIT Jodhpur 7

 Scaling In Geometry
If we let ℓ = linear dimension of a solid, we will have:

The volume: V  l3 S/V = ℓ-1

The surface:
S  l2 A 10 times reduction in length

→ 103 = 1000 time reduction in volume.

but → 102 = 100 time reduction
in surface area.
Since volume, V relates to mass and surface area, S relates to buoyancy force:

S/V ≈ 10-4/mm

S/V ≈ 10-1/mm

So, an elephant can never fly as easily as a dragonfly!!

24-Jun-20 IIT Jodhpur 8

Example : Example on scaling law in geometry of a device

What would happen to the required torque to turn a micro mirror with a
50% reduction in size?
y  I
Torque required to turn the mirror: 
where Iyy = mass moment of inertia of the mirror about
yy
y-axis determined by the following expression:
1 Mc
I yy  2
b
12
in which M = mass of the mirror and c = width of the
mirror.
But the mass of the mirror, M = ρ(bct) with ρ = mass density of the
c mirror material (a fixed value). Thus, we have:
1
y
I yy  bc3t
12

Meaning a factor of 32 times reduction in required torque to rotate the mirror!!

24-Jun-20 IIT Jodhpur 9
Scaling In Rigid Body Dynamics
 Forces are required to make parts to move such as in the case of actuators.

 Power is the source for the generation of forces.

 An engineer needs to resolve the following issues when dealing with the
design of a dynamics system such as an actuator

 The required amount of a force to move a part, How fast the desired movements
can be achieved, How readily a moving part can be stopped. The resolution to the
above issues is on the inertia of the actuating part.

 The inertia of solid is related to its mass and the acceleration that is
required to initiate or stop the motion of a solid device component.

 In the case of miniaturizing these components, one needs to understand the effect of
reduction in the size on the power (P), force (F) or pressure (p), and the time (t) required to
deliver the motion.

24-Jun-20 IIT Jodhpur 10

 Scaling In Rigid Body Dynamics
Rigid body dynamics is applied in the design of actuations and sensors, e.g.
accelerometers (inertia sensors).
It is important to know how size (scaling) affects the required forces (F), and thus power
(P) in the performances of these devices.
The dynamic force (F) acting on a rigid body in motion with acceleration (a) (or deceleration)
can be computed from Newton’s 2nd law: F = M a

The acceleration (a) in the Newton’s law can be expressed in the following way
In scaling:

Let the displacement of the rigid body, s  (ℓ), in which ℓ = linear scale. But velocity, v =

s/t, and hence v  (ℓ)t-1, in which t is the required time.

From particle kinematics, we have: 1
s  v ot  2a t 2
where vo = the initial velocity. 2s
By letting vo = 0, we may express: a
t2
Thus, the scaling of dynamic force, F is:
F  Ma  2sM
2  (l)( l ) t
3 2
t
24-Jun-20 IIT Jodhpur 11
 Trimmer Force Scaling Vector

24-Jun-20 IIT Jodhpur 12

 Scaling In Rigid Body Dynamics
Trimmer force scaling vector
 Power density (P/Vo):
 When scaling down a MEMS or a microsystem, one must make sure that the
power used to drive the device or system is properly scaled down too.

 In the design practice, power density, rather than power, is used.

 Power is defined as energy produced or spent by the device per unit time, and energy
is related to work, which is equals to the force required to move a mass by a distance.
Mathematically, these relationships can be expressed as:

P  Fs
Vo t Vo
in which, F = force, s = the displacement of the mass moved by the force, and t =
time during which the energy is produced or consumed.
 The above expression is used to derive the “Force-scaling vector” as
shown in the next slide.

24-Jun-20 IIT Jodhpur 13

 Scaling In Rigid Body Dynamics
Trimmer force scaling vector-

24-Jun-20 IIT Jodhpur 14

 Example
Estimate the associate changes in the acceleration (a) and the time (t) and
the power supply (P) to actuate a MEMS component if its weight is reduced
by a factor of 10.
Solution:

Since W  V (= (ℓ)3 , so it involves Order 3 scaling, from the table for

scaling of dynamic forces, we get:

 There will be no reduction in the acceleration (ℓ0).

 There will be (ℓ0.5 ) = (10)0.5 = 3.16 reduction in the time to complete the
motion.

 There will be (ℓ0.5) = 3.16 times reduction in power density (P/Vo).

The reduction of power consumption is 3.16 Vo. Since the volume of the
component is reduced by a factor of 10, the power consumption after
scaling down reduces by: P = 3.16/10 = 0.3 times.
24-Jun-20 IIT Jodhpur 15
 Scaling in Electrostatic Forces

24-Jun-20 IIT Jodhpur 16

 Scaling in Electrostatic Forces
Electrostatic forces in misaligned electrodes are obtained by:


2
U 1 WL V
F   0 r
 l2
Fd
d
 2 d
2

Fw d

L FL
F w   U  1
 o r V
L 2
 l2
V d d
W W

U

1 0 r
WV
2

  l2
FL  2L 2 d
So, we may conclude that electrostatic forces:

Fd, FW, and FL  l 2

Scaling: A 10 times reduction in electrode linear dimensions

→ 102 = 100 times reduction in the magnitude of the electrostatic forces.

24-Jun-20 IIT Jodhpur 17
 Scaling in Electromagnetic Forces

24-Jun-20 IIT Jodhpur 18

 Scaling in Electricity
Electric resistance R  L  l1
A
in which , L and A are respective electric resistivity of the material, the length and
across-sectional area of the conductor

Resistive power loss P  RV  l1

2

where V is the applied voltage.

U  1  E 2  l 2
Electric field energy 2
where  is the permeativity of dielectric , and E is the electric field strength  l1 .

Ratio of power loss to available power P  l1  2

E av l l
3

24-Jun-20 IIT Jodhpur 19

 Scaling in Electricity
From “Nanosystems,” K. Eric Drexler, John Wiley & Sons, Inc., New York,
1992 Chapter 2, ‘Classical Magnitudes and Scaling Laws,’ p. 34:

Electric Quantity Index, α in ℓα

Current, i 2

Voltage, V 1

Resistance, R -1

Capacitance, C 1

Inductance, L 1

Power, P 2

24-Jun-20 IIT Jodhpur 20

 Scaling In Fluid Mechanics

Flow in Flow out

Q (Effect of scaling
a
down the radius of
p tube i.e. Keeping
radius “a” in
consideration, having
fixed length L)
Volumetric Flow, Q:
From Hagen-Poiseuille’s equation:

Q
 a4 P Leads to: Q  a4 (Approximated)
8L

Meaning a reduction of 10 in conduit radius

→ 104 = 10000 times reduction in volumetric flow!

24-Jun-20 IIT Jodhpur 21

 Multi-scale Manufacturing
• Introduction:
 There is currently a worldwide effort for advances in micro and
nanotechnologies due to their high potential for technological applications in
fields such as micro electromechanical systems (MEMS), organic electronics
and high-performance structures for aerospace.

 Due to the several orders of magnitude involved in the fabrication of nano

composite devices, an efficient manufacturing technique must address the
challenges at the nano, micro and macro scales.

 The capability to manufacture complex 3D structures will greatly expand the

practical applications of nano composite materials and enable the development
of novel devices such as a 3D nano composite micro-coil spring.
 Multi-scale Manufacturing-Example
 Figure below shows this multiscale concept for the creation of a 3D scaffold
structure using a single-walled carbon nanotube (SWNT) and a polymer nano
composite.

 At the micro-scale, the production of nano composite structures must allow a

control over the orientation of high aspect-ratio nanoparticles such as SWNT.
 The arrangement of the nano composite micro-scale structures in 3D permits the
localization and orientation of the nano-reinforcement in a macro-scale product.
 Nanoscale
 The nano scale poses important manufacturing challenges such as the dispersion
•  of the nanoparticles and the close interaction of the nanoparticles with the host
polymer matrix.
 For mechanical reinforcement, every nanoparticle should be separated from each
other to maximize the interface between the matrix and the nano-reinforcement,
thus enhancing the available area for stress transfer to the nano-reinforcement.
 The maximum volume concentration (Vf,max) of cylindrical particles in a
hexagonal packed arrangement is expressed as a function of the ratio (r) of the
diameter of the cylinders (d) over the separation distance (c) between them.
 Vf,max= where r=d/c.

 The interface zone is a certain thickness around nanoparticles where the polymer
conformation is influenced by the nanoparticle. The value of the polymer radius
of gyration is commonly employed to gage this distance.
Nanoscale
 Microscale
Advantages over Nanoscale Manufacturing:
 The material needed is reduced when the cost is an issue.
 The microstructures manufacturing techniques allow a better control on the
nanoparticle disposition due to their microscale confinement.
 High aspect ratio nanoparticles, such as CNTs, can align themselves along the
flow direction with the help of the high shear achievable in small-scale
manufacturing. Several techniques are used to produce nano composite at the
microscale.
Example-fabrication Of Polyurethane
Nanocomposite Structures
 Example-Fabrication of polyurethane
nanocomposite structures
• Step 1:Nanoscale: integration of SWNTs to a polyurethane
polymer :
 Before integration to the polymer, a reflux of five hours in a nitric acid
solution is used to treat the as-grown SWNTs.
 A non-covalent modification is also performed by sonicating a specific
amount of the purified SWNTs .
 A commercially available UV-curable polyurethane is then slowly added to
the SWNT solution in DCM while stirring with a magnetic stirrer.
 After evaporation of the DCM solvent the nano composite mixture is passed
several times in a three-roll mixer mill where the gap between the rolls and
the speed of the apron roll are controlled.
 Fumed silica nanoparticles are also added to increase the viscosity and to
yield a shearthinning behaviour.
 Finally, the nanocomposite was poured into syringe barrels, where the DCM
solvent was fully evaporated.
 Example-Fabrication of polyurethane
nanocomposite structures
• Step 2:Microscale: UV-assisted direct-write fabrication of nano
composite fibers.
1. The UV-curable nano composite can be placed in a syringe barrel equipped with a micro-
nozzle tip. Under the application pressure, the gel-like nano composite is extruded
through the micro-nozzle to form a fiber.
2. Then, using a UV source that follows the extrusion point, the material is cured right after
extrusion to form a fiber that can span between two support points.
3. A fiber is extruded between two polymer pads by robotic displacement of an extrusion
nozzle. The fiber is exposed right after extrusion to a UV-source to cure the nano
composite.
 Example-Fabrication of polyurethane
nano composite structures
• Step 3:Macroscale: fabrication of 3D SWNT/polyurethane nano
composite structures using direct-write techniques:
 This approach permits the fabrication of a 3D-reinforced product through the
directed and localized infiltration of SWNT/polymer nano composites into a 3D
microfluidic network.
 The direct-write assembly method is first used to generate a scaffold of fugitive
ink filaments.
 The resin is then allowed to harden around the fugitive ink filaments. Heat
generation during curing reaction of the resin is kept to a minimum to prevent
deformation of the fugitive ink scaffold.
 The final step is to cure the nano composite to obtain an epoxy beam reinforced
by a 3D network of nano composite fibers. This technique allows the designer to
place nano composite filaments were they are needed in a macroscopic product.
Example-Fabrication of polyurethane nano
composite structures
Thank you