Parametric Modeling of the Human Body for Technical Aims
D. Cambiaghi, E. Chirone, S. Uberti, V. Villa, N.Martinelli
Dipartimento di Ingegneria Meccanica, Universit di Brescia
Via Branze, 38 - 25123 Brescia - Italy
e-mail: villa@ing.unibs.it
Keywords: parametric modeling, man machine interaction, human factors
1 AIMS AND OBJECTIVES
This work was born from the need to have a software tool which could help the designer to
evaluate the physical man-machine interaction in mechanical projects where the morphology
is the key parameter to develop user oriented machines.
IQUKE is a research project, lately started, aimed to develop a parametric model of the human
body.
The model will be developed in the CAD environment where it is to be used, while the
import/export of the model from/to other CAE/CAD environments must be avoided to prevent
loss of parametric and kinematics properties as well as undesired simplification.
2 CONTRIBUTION OF THE PAPER
2.1 Background
At the University of Brescia the MAR
1
group operates in developing devices to help disabled
people to improve their standard of living [2], [3]. At the same University operates the VDG
2
group which deals with race cars. They both have the need to use reliable models of the
human body, and they both have found the models available on the market unsatisfactory,
either due to the lack of flexibility of use and cost problems, or others.
2.2 Machines and devices heavily interfaced to the human body
The relevance of the man-machine interface has been analysed and several degrees have been
considered.
Some examples follow, as experienced by the authors.
1
Mechanics for Autonomy and Rehabilitation
2
Vehicle Dynamic Group
In a car the interface is generally well known, being basically composed with the driving
wheel, the seat, the pedals and the cockpit, which must be tuneable to fit a large range of
human sizes.
In a car for disabled people the interface must be more carefully tailored to the foreseen user.
In the case of a measuring arm spasticity device [5] the interface must be deeply detailed with
regard to the arm to be monitored, while it can be quite rough for the remaining of the body.
In the case of the wheelchair the whole wheelchair itself is both the machine and the interface.
Figure 1. Car Figure 3. Car for Disabled People
Figure 2. Measuring Arm Spasticity Device Figure 4. Wheelchair.
3 RESEARCH ABOUT EXISTING COMMERCIAL MANNEQUINS
A previous analysis of the state of the art showed that the following models are available:
1. surface models developed for animation purpose, graphically advanced but poorly
equipped with inertial and kinematics features;
2. CAE models developed for ergonomic analysis, quite expensive, requiring powerful
computers and/or long computing time, rigidly limited to their native environment, and
consequently poorly connectable to machine models developed elsewhere;
3. dedicated CAE models (for instance aimed to study the spinal bone) which are extremely
stiff in use;
4. FEM models, generally used for crash test simulation, useless for the aimed purpose;
5. other models, equally useless for various reasons.
In fact these products are stand-alone and they have their own database and software
environment so they dont give the necessary flexibility to help the designer in developing his
projects.
This products are not CAD in the strict sense of the term because they force the designer to
redraw the entire project in the environment of the human modeling software to carry out the
analyses he needs.
Usually such software arent user friendly in the drawing field such as the modern commercial
3D CADs are (SolidWorks, Think3, Inventor, Unigraphics, etc.). As a consequence they
arent useful for the average designer.
Furthermore these human modeling tools dont usually allow any customisation and
morphological control on the created models.
As a consequence it has been suggested that the development of a model fit to overtake the
abovesaid limitation may become the subject of a dedicated research program. Having found
interest in the academic area, the program started
Figure 5. ManneQuin Pro - NexGen Ergonomics Figure 6. Thorax FEM Model Figure 7. Animation Purpose Models
4 WORKING IN A CAD ENVIRONMENT
The necessity was recognised to develop a flexible and modular model available in the same
CAD, in which the interfacing machines are developed too.
We thought to proceed in an a way which can be considered opposite to the usual: the idea
was to study and realise an anthropometric model inside a traditional CAD environment, the
same environment where the machine models are developed.
Working in a CAD environment with perfectly known and controlled geometry offers some
advantages:
1. In every moment the morphology of a body segment or of the whole body can be handled
and modified.
2. We can choose some different anthropometric configurations and observe the effects of
the change on the projects
3. Positioning the model can be extremely precise.
4. It is possible to measure the angles between the body segments in the model.
5. A large number of CADs can be programmed with relatively simple tools, so appropriate
user interfaces to drive the human model can be created.
6. It is possible to carry out dynamic and structural analyses because usually CADs have in
them some tools to execute FEM or dynamic calculations.
5 MAIN TARGETS
The basic target was the definition of the general strategy for the implementation of the human
model in a generic CAD environment. The main problem is to evidence the theoretical bases
necessary to develop a model in every CAD which requires such model.
It is desired that the strategy ends in a model featured with:
1. good morphology, with regard to encumbrance, aesthetical aspect, graphic rendering;
2. kinematics of joint, with regard to shoulder, ankle, hip, knee, spinal bone, and so on;
3. kinematics of joint, with regard to of the joint Range of Motion (ROM);
4. parametric structure, with regard to the whole body and related to height, mass, age, sex,
ethnic origin and so on;
5. mass properties, with regard to mass, COG location and inertial moments of each
segment;
6. exportability, with regard to whatever CAD environment;
7. modularity, with regard to possibility to integrate in the same model quite simple and quite
complex segments/joints.
Figure 8. Joint Range of Motion
6 DEVELOPMENT OF THE STRATEGY
The first step was the collection of the human body anatomic and anthropometric data with an
adequate detail. In fact we thought to realise a simplified model with the required properties
but without the extreme complexity of the human skeleton. We discovered that, accepting a
few percentage of inaccuracy, some simplified joints could be used with success and it was
also possible to use some simplified segment morphologies, each driven parametrically by few
dimensions (fig. 9).
It is here to be pointed out that the modularity of the software will allow at any time to
increase (or to reduce) this level of detail.
So, after that analysis we decided to start with 17 body segments with 40 degrees of freedom.
After a research, some accurate databases were found: even if they are not representative of
the entire human population they are useful for this step of development.
In fact these databases are built upon a military population by USAF and NASA for their
specific and unusual purposes, they are very accurate, in particular about the COGs and the
inertia moment of the bodies analysed; this was very useful to build and validate our model.
The earth of the strategy lies on two basic solid models (male and female, fig 10) associated
with a high number of dimensional tables for each database; each body segment is associated
with more tables (one for each database) containing the values of the characteristic
dimensions of that body segment: working on those values (choosing for example the values
associated with a determined percentile or changing the whole table) we can configure the
body.
In the future this will allow to build (having the morphological data to fill the table) fat or slim
people or people with short legs or long arms and so on.
We can also represent every human being tailoring the model on him, inserting his body
measures in the tables of every body segment.
All this procedure is managed by a program with a simple interface (programmed in Visual
Basic) running under the CAD environment.
The model is built upon the basic dimensions of each body segment which can be extracted
from the databases previously quoted.
Figure 9. Parameterised dimensions of a Body Segment (Leg).
The program uses the interfaces which the CAD (SolidWorks) gives to the user (in this case
the API - Application Program Interface) and it commands directly the drawing functions of
the CAD.
An important feature is the fact that, with this sort of strategy, the databases can be changed or
integrated in every moment with some simple operations.
This feature will allow to extend the validation towards the edges of the Gaussian distribution
of every human characteristic (weight, body configuration, morphological race, age) having
the appropriate database.
Figure 10. Basic Models.
6.1 Using E-NES
Now well explain a typical cycle of validation of a man-machine interface using E-NES
which is the name given to the first running software of the IQUKE project.
At first the designer launches the software and chooses a configuration for the model; so the
software generates the model.
s se ex x: : M Ma an n
p pe er rc ce en nt ti il le e 5 50 0 ( (1 17 75 5 c cm m) )
d da at ta ab ba as se e: : E Eu ur ro op pe e
b bu ui il ld d A Av ve er ra ag ge e
w we ei ig gh ht t 7 74 4 K Kg g
Generation of the model Loading the position from file (under development)
s se ex x: : W Wo om ma an n
p pe er rc ce en nt ti il le e: : 5 50 0 ( (1 16 68 8 c cm m) )
d da at ta ab ba as se e: : E Eu ur ro op pe e
b bu ui il ld d: : S Sl li im m
w we ei ig gh ht t: : 5 52 2 K Kg g
Changing anthropometric parameters
Figure 11. Using E-NES.
Now we can verify our project by changing the morphology of the user or analysing the model
with the tools of the software (under development) or of the CAD.
Figure 12. Using E-NES.
The designer can handle the body of the model like all others mechanical object built in the
CAD, for example to draw a race car seat around it.
6.2 Movement control
The part of the software dedicated to the movement control is currently still under
development.
In a first time we thought to use some specific surfaces applied to the body segments to drive
and control the body movements using the CAD tools (e.g. interference controls, etc.), but it
revealed to be too complex and it made the computer slow his speed which is in contrast with
our goals.
So we decided to use the techniques related with the Robotic Mechanics to control the
movements. In fact using the API it's possible to read the Position Matrices of each body
segment handling them with routines realised by us or using DLLs developed by the robotic
work group of the University of Brescia (Prof. G. Legnani).
When it will be completed we will be able to simulate each positioning and create some
sequences of movement like walking for example.
It will also be possible to read in output positioning file and verify every angle or position
with appropriate databases from medical literature to evaluate, for example, the comfort level
of a particular posture, etc.
7 RELATED APPLICATION AND FUTURE DEVELOPMENTS
7.1 Related Application
1. tool to help designing machines deeply conditioned by the man/machine interface (cars,
devices to help disabled peoples, rehabilitation);
2. safety and ergonomics;
3. solid models for FEM analyses (crash test, for instance).
7.2 Firsts developments
We lately start a collaboration with Laben S.p.A. (a Finmeccanica company) based in
Vimodrone (Milan) which works in the aerospace field.
The contract goal is to develop a parametric model with an improved detail of neck and head.
Our models will be used in the ALTEA program (Anomalous Long Term Effects on
Astronauts) funded by the Italian Space Agency (ASI) and entirely under Laben responsibility.
The sophisticated equipment will be carried on board the International Space Station in the
US laboratory (Destiny) at the beginning of 2003 and then moved to the Russian module.
The purpose of this experiment is to investigate the origin of light flashes seen by astronauts
after a period of dark adaptation.
Over time, the scientific study has been extended to an evaluation of the effects of high-energy
particles on the brain.
As shown in figure 12, the astronaut under test has to wear a heavy helmet with high-energy
particle detectors (white boxes). In microgravity environment, weight does not represent a
problem but all the effects of the inertial properties have to be previously and accurately
evaluated in order to avoid consequences on the astronaut neck and head.
Figure 12. A.L.T.E.A Helmet (courtesy of Laben S.p.A.).
7.3 Possible future applications
1. Models of knee and spinal bone aimed to advanced ergonomic analyses (posture, weight
support, prosthesis impact, etc.);
2. muscle modeling (prosthesis design, clothing, etc.);
3. hand and foot modeling (sport, shoe design, etc.);
4. skeleton modeling (medical area);
5. further possibilities to tailor the model to unusual cases.
8 REFERENCES
[1] Martinelli N. Modellazione del corpo umano per usi tecnici con le tecniche del CAD
parametrico, Degree thesis, Universit di Brescia, A.A. 1999/2000.
[2] Cambiaghi D., Chirone E., Gadola M., Villa V., Solid modeling and virtual simulation
as tools to design for disability recovery, 12
th
ICED, Munich, Germany, 1999.
[3] Uberti S., Servomeccanismo per la salita in auto del disabile su carrozzina: studio,
sviluppo e progetto del dispositivo di aggancio, Degree thesis, Universit di Brescia,
A.A. 1997/98.
[4] Winter D.A., Biomechanics and motor control of human movement, Waterloo, CND.,
Wiley, 1990
[5] Cambiaghi D., Cattaruzzi L., Malgrati D., Villa V., "Studio di un dispositivo
elettromeccanico per la misurazione della "spasticit" nel paziente affetto da danno
neurologico centale", Brescia Ricerche - Pubblicazione INN.TEC. - Anno XI - N. 32 -
Settembre 2000 - Brescia, Italy
