You are on page 1of 61

Risk quad chart NEW Testing 2 types of the punch-modular design

Trimming
Pareto DFMEA

Design of Experiment Axiomatic Design

ARIZ
QFD NEW Rapid 3D prototyping

Tree diagram
Trace diagram
NEW Capability and Design scorecards
NEW Videos of punch while cutting with the springs Flow charter
40 invenitve principles
The DFSS project, designed with the Acclaro DFSS, Minitab, CATIA, TRIZ and Six Sigma
Rajko Šrimpf, Black Belt Six Sigma, Design for Six Sigma, Applied Reliability, Triz Level 3, March 2014
IDENTIFY
Project name: Unique paper hole punch

Project overview:

The scope of this project is to design new type of paper hole punch like nothing else in the world market. The
new design is going to be lightweight and less expensive than competing product of this type on the market.

Problem statement:

Current puncherers are big and over designed with material because of the strong springs witch are needed to
open the puncherer. Our design will save material and production cost. The goal is to use as little production
equipment as possible to meet our cost goals.

Customers/Stakeholders:

The new design would appeal to the home and office costumers, who don‘t need a standard big puncherer and
can go for a cheap alternative that can still do the job.
Goal of the Project:

To design the cheapest possible hole punch, that can still do the job against competing products. We know that
some equipment sells because of design appeal. We are sure that for a non-professional customers all that
matters is the cheap price and workable puncherer.

Scope Statement:

I intend to design, and make model with the 3D printing to prove whatever design will come out.

Project financial benefits:

This is so called the project to test DFSS and TRIZ tools, and there is no better way than to learn from a real
thing. This is so called TRIZ level 3 certification project.
Maped, the smallest one No name, small

Laco model L350, medium Leitz 5008, medium


Some old stuff:
Puncher is 3 Stage product

We hit developement limits.


Products sell on design apeal and price.
Recomendation: COST REDUCTION!
But how? Everbody looks almost the
same and uses the same solutions.
Can we find another?
DEFINE
We will use TRIZ tools here
Six Sigma DOE here

Pie Chart of CTQ

Category
8,9% 9,8% Remov able Pad
New body of the punch
Dev ice to adjust paper size
Proper angle and size of the lev er
Special cutting tools

32,1%

41,1%

8,0%
Red line is a loop!
DESIGN
We will trimm this holders.
Somehow we have to add this
features to the pad of the
puncher.

This is the pad. How on Earth can


we add holders to the pad?
We have added all holding function to the Unibody. But how can that be? It has to be revolutionised in
every way.
Spring and cutting tool holders

Paper removal
Axl holes. To set them at optimum
position we will do Design of
Experiment

Now we have the pad. What about the


stability of the Unibody? We have
another TRIZ problem. We will use ARIZ
to solve this problem.

We added grooves for stability


Angle The problem: Where to set axl against the rod?
L
We have x,y coordinates as Factors and two Responses L and Angle!
Pareto Chart of the Standardized Effects Pareto Chart of the Standardized Effects
(response is Angle, Alpha = 0,05) (response is L, Alpha = 0,05)

2,57 2,78
F actor N ame
A X
B Y

AB
X

Term
Term
B

Y
Angle A L
0 10 20 30 40 50 60 0 2 4 6 8 10 12
Standardized Effect Standardized Effect

Region of the optimal set-ups in white


L distance factors importance Contour Plot of L vs Y; X Contour Plot of Angle; L

5 5
Category Angle
Ostalo X L
2,9% X 20
8,1% < 14,2
Y 30
14,2 – 14,4
Y X*Y
8,1% 4 14,4 – 14,6 4 L
Ostalo 14
14,6 – 14,8
14,8 – 15,0 17
15,0 – 15,2
3 15,2 – 15,4 3
> 15,4

Y
Y

2 2
X = 19,7774
Y = 2,52626
Angle = 25,0147
1 L = 14,4038
1

X*Y 0 0
80,9% 10 12 14 16 18 20 10 12 14 16 18 20
X X

This is very unexpected. I newer had DOE with such a powerful interactions!
If we use rings only, than we
have problem with the stability
of the Unibody! Ring compresses the spring
EC-1: If we use holders and don‘t use Unibody, then the punch is stable, but we consume to much of
material.
EC-2: If we use Unibody, then the punch is not stable, but we consume less material.
IFR: X-component eliminates the harmful action of unstable design and
Product: Punch allows to deliver the main function of the system this is stability of the
Tool: Unibody punch in the operating space that is the same in the punch during the
operating time, which is the same without making the system more
Graphical representation: complex and without any harmful consequences.

EC-1 EC-2

Resources: air, shape of the rings, hands, table, cutting tools, springs, shape of the lever, shape of the
axl, rod, pressure

Solution of the problem:

Resource parameter Shape of the rings should be used to connect the cutting tools for stability.
Resource parameter Shape of the axl should be used to connect the part of Unibody for stability.
Resource parameter Shape of the rings should be used to connect the cutting tools for stability

Profile takes the role of the ring and


compresses the springs and holds
with the cutting tool unibody locked Movable profile

Fixed profile

Axl doesn‘t have only rotation


function, but also stabilizes the Couldn‘t have keeped it for myself, but those two
Unibody. But for how much? Is there profile auto-aligne the cutting tools if you try to
a way to extend its stabilization? But bend it. They will allways go to the start position.
how? Of course the ARIZ is on the template that‘s 8 pages long. This Tested in on prototype, see the movies.
is only the solution of the problem.
Resource parameter Shape of the axl should be used to connect the part of Unibody for stability.

Place one object inside of another


With the lever and O profile in
the position, axl slides throught
them and then we deform the
tips of the axl to prevent falling
out of the punch.
O profile Additional surface for contact

Axl

The axl and O profile locks thight the Unibody.

We have increased the surface contact between axl + O profile and Unibody.
Pressure Tension

The problem is the bend on the outside like here on the


Profile adds a lot of stability to the Unibody. You can not picture. We have solved this problem with the lever,
bend it inside. which covers the outside and prevent bending of the
Unibody. The tolerance between the Unibody and the
lever has to be optimal to allow roation of the lever and
not allowing bending unibody on the outside.

This gap is to big in our model. By my calculation it should be


about small to prevent from bending and allowing to move the
lever easily over the Unibody.
While deforming into the groove, material gets harder around it and in it, thus prevents bending.
While deforming into the groove, material gets harder around it and in it, thus prevents bending while applaying force on
the lever, hence while cutting the paper.
EC-1: If the pad is big in Unibody, then we can cut the paper with no problems, but we use to much
of the material.
EC-2: If the pad is small in Unibody, then we can not cut the paper with no problems, puncher will
flip over, but we use less material.
IFR: X-component eliminates the harmful unstability of the punch and
allows to deliver the main function of the system this is stability in the
Product: Punch
operating space that is the same during the operating time, which is
Tool: Pad
not the same (only when cutting) without making the system more
complex and without any harmful consequences.
Graphical representation:

EC-1 EC-2
small
big

Resources: air, profiles, hands, table, cutting tools, springs, lever, axl, rod, pressure

Solution of the problem:

Resource parameter Hand should be used to press on the pad for stability while cutting the paper. Of
course is the ARIZ on template that‘s 7 pages long. This is only the solution of the problem.
With second hand we press on the
lever of the pad to cut the paper
With one hand we press on the
backside of the pad Inventive principles for seperation in time:
Principle 9: Preliminary anti-action
Principle 10: Preliminary action
Principle 11: In-advance „cushioning“
Principle 15: Dynamics
Principle 34: Discarding and recovering

Since we have separation in time we look for the principles that solve this problem. The principle 9:
Preliminary anti-action was used. If it is necessary to perform an action with both harmul and useful effects,
this action should be replaced with anti-action to control harmful effects.
Red line is a loop!
This rotation actually allowes to use the second hand!!!
Adds stability while cutting

Space for the paper clips


Type 2 pad optimized with increased Type 3 pad optimized with increased
Type 1 pad space for paper clips space for paper clips without one support

With type 2 and 3 pad we end up with more paper clips space and allow the Unobody main pocket
to be deeper thus preventing ther Unibody to bend in any way while cutting.
First model Added grooves Added chamfer Added pocket

Increased size of the pocket Pocket is deeper

Added big
pocket

Added bend of material for stability Bend removed


I have made a simplification of the steps needed. If it is necesary we can add
more steps espeacially at cutting. This is only represantation, actual space
between stations might vary, let‘s say they are probably bigger. I think we
have to add calibration step for bend parts of the product . What is
Axiomatic Design saying about the complexity of the steps in the press tool?

Principle 10:
Preliminary
action, we
drew pocket
deep before
cutting.

If there is a
problem with this
pocket when Cutting holes!
cutting we could
me it a bit smaller! We need to separate steps so
Like this. there needs to be cutting tool or
we cut after the pocket!
Only the lever is different between this two
types of products. A modular design, who
tought about that in this industry?

Type 2: Big lever

Type 1: Small lever


Axiom 1: The Independnce Axiom:
Maintain the independance of the Probability of design
functional requirements. Category
Uncoupled design Uncoupled design
6,3% Decoupled design
Coupled design

Axiom 2: The Information Axiom:


Minimize the information contect in
a design.
Decoupled design
37,5%
Coupled design
56,3%

So 56,3 % of all designs are coupled design? Is it just


me to see it that way, or this is a major problem for
manufacturers?

We have decoupled design, so it is than more likely that we will achive Six Sigma quality level. So, I
would be happy even with the 5 Sigma level.
6,83 is left for the hand that does 26 33,29 is left for the hand that does 26
- Less expensive, cos‘ less machinery
and people manufacturing it
- More portable, less weight (big
punch)
- Design appeal

- Cutting competitive amount of


paper
- Paper size adjusterer

- Have to use both hands only in the big one


- Less capacity for paper clips, oh, this is not a
problem as I see it. Who colects clips anyway?
Occurrence

Detection
Severity
Detection

RPN
Item Function Failure Mode Failure Effect Potential Failure Cause Preventive Action
Action

Cutting tool Cutting Broken tool Not cutting properlly 1 Bad material Proper material 1 Visual 1 1

Profile Compressing the spring & stability Broken Not compressing the spring 2 Bad material Proper material 2 Not working 1 4

Spring To lift the tool Broken Not working 1 Bad material Porper material 1 Not working 1 1

Profile Stability Bend Failled stability of the unibody 2 Bad material Porper material 2 Not cutting 1 4

Pareto Chart of Item


Unibody Holding the cutter together Bend unibody Not cutting properlly 3 Failled profile Quality sheet metal tool 2 Not working 1 6
25 100
Lever Axial movement Bend Not working 1 Failed Perpendicularity Quality sheet metal tool 1 Not working 2 2
20 80
Base Holding the Unibody Doesn't fit Can not attach to unibody 1 Bad plastic tool Adjust tool properlly 1 Doesnt' fit 1 1

Percent
15 60
RPN

P. adjusterer Adjusting to the paper size Doesn't fit Can not adjust paper size 2 Bad plastic tool Adjust tool properlly 1 Doesnt' fit 1 2
10 40
Axl Rotating the lever Broken Not working 1 Bad plastic tool Adjust tool properlly 1 Doesnt' fit 2 2
5 20
O profile Adss stability Broken No stability of Unibody 2 Bad tolerance Proper tolerance 1 Doesnt' fit 1 2
0 0
Item l r er ol r
ile dy Ax ve ile se he
rof bo Le of te
r
Ba to
Ot
P i pr s g
Un O ju ttin
ad Cu
P.
RPN 8 6 2 2 2 2 1 1 1
Percent 32,0 24,0 8,0 8,0 8,0 8,0 4,0 4,0 4,0
Cum % 32,0 56,0 64,0 72,0 80,0 88,0 92,0 96,0 100,0
OPTIMIZE
3D RAPID PROTOTYPING
So, here are the final dimensions of the holes. With the 3D print they will vary. Will do scorecrads for each of them, but
only for 4 or 5, cos I will not produce then just makeing the prototypes. Of course, if you are reading this you might be
Six Sigma expert and understand the importance of low variation of the dimensions of the parts to be assembled
togehter.
Since I am gonna print many cutting tools for tests, I will make capability analysis in Minitab to see
how stable is the printed process for sample size.
I had two suppliers for this prototype. To bad in Slovenia they
just couldn‘t make them. Very disapointing to say at least. I have
choosen 3DShub and have found a supplier in Austria. He did a
very nice job. The pad rubber was printend in Belgium. So I have
international suppliers. This prototype is partly functional, I still
have to make rods, profiles, cutting tools and some cheap
springs. I bet it will cut one sheet, just to show that the design
works. Black rubber looks good, will use it in the future.
Kinematics is now fully optimizied, even the pitch and the hight of the springs are now adjusted to the rotation of the lever
and moving of the cutting tools. It took some time, to bad you can not do it with the one update in Catia, but you have to
upadate the model twice. No exotic kinematics add-ons were used.
All measurement in mm .00 All measurement in mm .00

3,95

5,60

?
3,96
4,61
Cutting holes
4,95

Can‘t belive they almost messed up the rubber part. It‘s wieder and the paper adjuterer won‘t fit in. What the hell?
But support for the Unibody is perfect. Couldn‘t have done it better. I have to speak with the supplier. Can not allow
the parts for the testing to be missing the original specifications. The Unibody part is printed by far the best with the
new supplier, a bit more expensive, but top quality. Shame on you rubber guys. How to measure holes that can not
be measured cos‘ of collision with the part with my brand new Holex? Read the next slide!
Gage Run Chart of Meauserument by Part, Operator
Gage R&R (ANOVA) for Meauserument Reported by :
G age name: Tolerance:
Reported by : Rajko Š rimpf D ate of study : M isc:
G age name: G age diameter of the cutting tools Tolerance:
M isc:
1 2 3 4 5 O perator
Components of Variation Meauserument by Part 4,475 A
B
100 % Contribution 4,48 C
Mean 4,450
% Study Var
Percent

% Tolerance
4,44

Meauserument
50 4,425

4,400
4,40 6 7 8 9 10
0
Gage R&R Repeat Reprod Part-to-Part 1 2 3 4 5 6 7 8 9 10 4,475
Part
R Chart by Operator 4,450 Mean
Meauserument by Operator
A B C
Sample Range

0,02 UCL=0,02145 4,48 4,425

_
0,01 R=0,00833 4,400
4,44
Operator
0,00 LCL=0
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Panel variable: Part
4,40
Part
A B C

Xbar Chart by Operator


Operator
You se small variation between mearusement? I
A B C Part * Operator Interaction
4,48
bet you won‘t belive the scale. It‘s 0.01 mm.
Sample Mean

UCL=4,45030 4,48 Operator


_
_ A
4,44 X=4,44178
Belive it or not. We did it. Took use 3 trials for
Average

B
4,44 C
LCL=4,43325
4,40
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
4,40
Gage from 90 % to 50 % and now 26,35%. This
Part 1 2 3 4 5 6
Part
7 8 9 10
is as good as it gets on solid sintering parts.

So, Gage R&R is marginal, study is 26,35% variance for the meausering system. Variance for the tolerance is 16,87 %. It‘s not
quite the best study we have ever did, but I am now confident that the meausering parts for system capability will be still
tolerable. Still Holex brand digital caliper si good, but Mitutoyos are the best, but still the 0.001 tolerance on sintering parts?
Process Capability Sixpack of Small So, the tolerance is within 0.2 mm.
I Chart Capability Histogram cpk is 1,73, what is translating to
4,50
UCL=4,5110 LSL USL

S pecifications
more than 5,2 Sigma. Long term
Individual Value

_
LS L 4,35 Sigma is 4,32 and ppm 7,83. Seems
X=4,4578 U S L 4,55
4,45 like 3D printing of some round
features is for the assembley of 3D
4,40 LCL=4,4046
1 3 5 7 9 11 13 15 17 19 21 23 4,35 4,38 4,41 4,44 4,47 4,50 4,53 printed punch enough. Didn‘t
Moving Range Chart Normal Prob Plot expected such a high Sigma. Had
A D: 0,367, P : 0,402
UCL=0,06535 more than 6 Sigma in some
Moving Range

0,050
previous project but this is really
0,025 __
MR=0,02
good news. All the cutters will fit
into the holes. There might be some
0,000 LCL=0
1 3 5 7 9 11 13 15 17 19 21 23 4,40 4,44 4,48 4,52 problems on the tip of the cutter,
Last 23 Observations Capability Plot that is usually bent on outside, but
4,52
StDev
Within
0,01773
Within
StDev
Overall
0,02131
you can put it togehter. Now all we
Z.Bench 5,20 Z.Bench 4,32 need is some cheap customized
Values

4,48
Cpk 1,73 O v erall Ppk 1,44

4,44
PPM 0,10 Cpm
PPM
*
7,83
springs…
S pecs
5 10 15 20
Observation
Process Capability Sixpack of Big Sigma is not as high as on the small
I Chart Capability Histogram diameter cos‘ the lower tolerance is
5,52
UCL=5,4986
LSL USL

S pecifications
out but it still long term Sigma is 3.73
Individual Value

5,46
_
LS L 5,35 and Ppk is 1.24, ppm is 95,24. This
U S L 5,55
5,40
X=5,4261 was in the line of expectation. The
LCL=5,3536
process is slightly moved to the left.
1 3 5 7 9 11 13 15 17 19 21 23 5,37 5,40 5,43 5,46 5,49 5,52 5,55 What I wanted to show is how Six
Moving Range Chart Normal Prob Plot Sigma is important on tolerance
0,10 A D: 0,724, P : 0,051
UCL=0,0891 calculations. Every Design for Six
Moving Range

Sigma project should have many Six


0,05
__
MR=0,0273
Sigma projects and of course be
0,00 LCL=0
finished with the Applied Reliability
1 3 5 7 9 11 13 15 17 19 21 23 5,35 5,40 5,45 5,50 and Lean for manufacturing. The p
Last 23 Observations Capability Plot value is on the limit, but it‘s above
StDev
Within
0,02418
Within
StDev
Overall
0,02039
0.05, so I did the capability.
5,450
Values

Z.Bench 3,15 Z.Bench 3,73


Cpk 1,05 O v erall Ppk 1,24
5,425 PPM 825,05 Cpm *
PPM 95,24
5,400
S pecs
5 10 15 20
Observation
I did the prototype with the springs. The design works perfectly, it even cuts the paper with the sintering cutting tools.
Who knew? It just works if you set your mind, time and limited resources to good use. It‘s like 6 Sigma. Did it allways
with no resources available and was sucessful most of the time.
So punch is working, little bends, but with the sheet metal they will be faded away… click on video!
Punching the holes…. Have to take into consideration that I am useing not metal cutters, but sintering, which
are not that sharp.
Punching the holes with the big one...
VALIDATE
And of course, not for this project but for solving ARIZ problem twice, I was rewarded with the TRIZ LVL 3 certificate!

I couldn‘t have helped myself for not posting TRIZ LVL 1 and 2 certificates also.
And I am also in the data-bank on www.matriz.org

You might also like