Mechanical Engineering Conference Presentations,

Mechanical Engineering
Papers, and Proceedings

9-2002

Modeling of Hydraulic Hose Paths

Kurt A. Chipperfield
Iowa State University

Judy M. Vance
Iowa State University, [email protected]

Follow this and additional works at: http://lib.dr.iastate.edu/me_conf

Part of the Computer-Aided Engineering and Design Commons

Recommended Citation
Chipperfield, Kurt A. and Vance, Judy M., "Modeling of Hydraulic Hose Paths" (2002). Mechanical Engineering Conference
Presentations, Papers, and Proceedings. 29.
http://lib.dr.iastate.edu/me_conf/29

This Conference Proceeding is brought to you for free and open access by the Mechanical Engineering at Iowa State University Digital Repository. It
has been accepted for inclusion in Mechanical Engineering Conference Presentations, Papers, and Proceedings by an authorized administrator of Iowa
State University Digital Repository. For more information, please contact [email protected].
Modeling of Hydraulic Hose Paths
Abstract
Hydraulic hoses are key components used to transfer power in heavy industrial machinery. The routing of
these hoses is currently performed late in the product design process because no accurate physical models of
the hoses exist that allow designers to predict the path the hoses will follow when installed in the machine.
Designers must either guess the path the hose will take based on prior experience or wait until the first
product prototype is built in order to experiment with the hose routes. This paper describes the use of
ADAMS, a commercially available dynamic modeling package, to predict hose paths. The hose path model
was verified by comparing the predicted paths to the paths of real hoses.

Keywords
VRAC, Hydraulic hose, Modeling

Disciplines
Computer-Aided Engineering and Design

This conference proceeding is available at Iowa State University Digital Repository: http://lib.dr.iastate.edu/me_conf/29
DAC TOC
Proceedings of DETC02
Proceedings of DETC02
ASME 2002 Design Engineering Technical Conferences
ASME 2002 Design Engineering Technical Conferences and Computers and
and Computer and Information in Engineering Conference
Information in Engineering Conference
Montreal, Canada, September 29-October 2, 2002
Montreal, Canada, September 29-October 2, 2002

DETC2002/DAC-34152

DETC2002/DAC-34152

MODELING OF HYDRAULIC HOSE PATHS

Kurt A. Chipperfield
1
Judy M. Vance
Department of Mechanical Engineering
Virtual Reality Applications Center
Iowa State University
Ames, Iowa 50011
[email protected]
[email protected]

ABSTRACT knowledge of the hoses minimum bend radii and flexibility.

Hydraulic hoses are key components used to transfer Later, during the prototype build, the hose lengths are adjusted
power in heavy industrial machinery. The routing of these using a cut-to- fit approach. Typically, one-fourth to one-third
hoses is currently performed late in the product design process of the hose lengths must be adjusted during the prototype build.
because no accurate physical models of the hoses exist that Hose clamp locations are identified, based on past experience
allow designers to predict the path the hoses will follow when and convenient attachment points, and the hose lengths are
installed in the machine. Designers must either guess the path measured to produce the manufacturing specifications.
the hose will take based on prior experience or wait until the
first product prototype is built in order to experiment with the Relying on past experience and the cut-to-fit approach
hose routes. This paper describes the use of ADAMS, a greatly reduces the ability to integrate hose design with product
commercially available dynamic modeling package, to predict design. Designers generally leave enough room for hoses and
hose paths. The hose path model was verified by comparing the hose specialists must route hoses through the space that is set
predicted paths to the paths of real hoses. aside for the hoses. Our approach models the physical paths of
the hoses so that the paths can be accurately predetermined
INTRODUCTION without relying on cut-to-fit. This allows the designers to
Hydraulic hoses, which are used to transport high-pressure design areas for hose routes as they are designing the product
hydraulic fluid, are key components in many large industrial and allows accurate hose specifications to be determined before
vehicles. A typical industrial backhoe machine contains the first prototypes. This will result in better designs, less hose
approximately 100 individual hoses. These hoses are routed failures, and faster time-to-market for the product.
around and through structural components of the backhoe and
present a challenge to product designers. The hoses vary in This paper presents the development of a hose model for
stiffness from almost rigid to very flexible and change shape predicting the shape of hydraulic hoses. The ADAMS [1]
when pressurized. Incorrectly designed hose routes can result in dynamic simulation software is used to calculate the predicted
hoses that rub causing failure. When designing such a vehicle, hose shape and a C++ software program provides a pre- and
designers must determine the hose routes and specify the hose post-processing interface. The hose model was verified by
lengths for part manufacture. comparing the output from the model with the path of a real
hose.
Currently, designers use CAD programs and rely on past
experience to plan hose routes. Hose specialists work with SUMMARY OF RELATED WORK
product designers to plan the hose routes and allow sufficient Several commercial CAD packages offer special modules
space for the hoses in a new vehic le. The hose routes are for the design of hose routes. For example, SDRC offers an I-
drawn into the CAD assemblies based on the hose specialists DEAS module for the design of wiring harnesses, cables, and
1
Corresponding Author

1 Copyright 2002 by ASME

hoses called I-DEAS Harness DesignTM. Parametric the same length and properties would take under the same
Technology provides a Pro/Engineer module called Pro/Piping. clamping conditions. This path is then returned to the hose
These packages allow the hose route to be easily drawn in the designer who considers if the hose route needs to be modified.
CAD environment and make the specification of hose ends
convenient but do not take into account hose properties such as Figure 2 shows a hose path as specified in the C_Hose pre-
weight, axial stiffness, torsional stiffness, or bending stiffness. processor (Fig. 2(a)) and the final hose shape as calculated by
Because of this, a hose path designed with these tools may not ADAMS Solver (Fig. 2(b)). The hose in the figure has fixed
take the same path as a physical hose. ends, no clamps, and lies in the x-y plane with gravity acting on
the hose in the negative y direction. The output of C_Hose is
Andrew Elliott [2] used ADAMS to model brake hoses and the positions of the points along the hose shown as large dots in
more recently, Rodriguez and Keil [3] published a methodology Fig. 2(b).
that also employs ADAMS to model hoses. ADAMS is a
multi-purpose mechanical simulation program sold by INPUT TO HOSE-MODELING PROGRAM: C_HOSE
Mechanical Dynamics, Incorporated [1]. The previous The input to C_Hose is a file containing a description of
ADAMS hose models were only developed to model hoses the initial hose path, hose properties, and clamping constraints.
whose paths are initially straight lines. The hose model The initial hose path is described by a list of hose point
presented here includes the ability to model complex hose coordinates, starting at one end of the hose and proceeding to
paths. In order to conveniently model these paths through the other. Hose points are illustrated in Fig. 2 as large dots.
industrial machines, a pre- and post- processor for ADAMS Also included in the input file are the physical properties of the
was developed. hose, the percent change in length, and the locations of any
clamps on the hose.
HOSE MODELING PROGRAM OVERVIEW
ADAMS is a general-purpose simulation package. An
ADAMS model may be built either by using the graphical user
interface called ADAMS View, or by creating a batch process.
Batch processing is performed using the ADAMS Solver
module. It would be time consuming to create complex hoses in
ADAMS View, so a separate C++ program was written in order
to conveniently pre- and post-process data to be used by
ADAMS Solver. This hose-modeling program will be referred
to as C_Hose. Figure 1 shows the organization of the C_Hose
program.

C_Hose
program ADAMS
Hose input file
input file y
Pre-processor

(a)
ADAMS x
Solver
system call

Post-processor
ADAMS
Hose output file
output file

Figure 1. Flowchart of the C_Hose program

USE OF HOSE-MODELING PROGRAM: C_HOSE

A user of C_Hose starts with an initial guess for the path a
hose might take through the machine. The user describes this
path to C_Hose by creating an input file that lists the x -, y -, y
and z - coordinates of points along the initial hose path from
start to end. The initial path may be designed using any CAD (b)
package or other design tool. The user also inputs the x
properties of the hose and which hose points are clamped.
After this information is provided, C_Hose calculates the "true" Figure 2. Example input (a) and output (b) from hose-modeling
path a hose of that length would take using ADAMS Solver. program.
The "true" path of a hose is the path a real, physical hose with

2 Copyright 2002 by ASME

 The required physical properties of the hose include its coordinate system. Usually there would be two markers per
weight per unit length, inside diameter, outside diameter, axial hose section (one at each end) but C_Hose allows each straight
stiffness, torsional stiffness, and bending stiffness. Percent hose section to be modeled with several equally spaced markers
change in length is the amount that the user wants the hose to in order to improve the accuracy of the solution. The number
shrink or expand from its original length. This property is used of markers per hose section is specified before a hose is
primarily for modeling pressurized hoses because most hoses modeled and remains the same for all hose sections. Markers
change in length when they are pressurized. The properties of a must be defined at the beginning and end of each beam
hose are not typically available from hose manufacturers so segment. If the number of markers per hose section were equal
they must be determined experimentally. See Chipperfield and to five, each hose section would be modeled as shown in Fig. 4.
Vance [4] for a description of hose testing methods. In this example there are four beam elements per section.
Markers 1-5 are on the first hose section and markers 6-10 are
Clamps give the user the ability to lock the position of the on the second section. Note that two markers coincide at each
input hose points. A clamp is specified at a hose point if the interior hose point: the marker at the end of the previous hose
hose must pass through that point. For example, a clamp would section and the marker at the beginning of the next hose
be used at a point in the route where the hose must be threaded section. Only one interior hose point is illustrated in Fig. 4
through a hole in a machine frame. Clamps are often used to because the hose shown is only composed of three hose points.
keep hoses from rubbing other components in the machine
while in service, which greatly reduces the life of the hose and 5
leads to hose failures.

C_HOSE PREPROCESSOR
The pre-processor part of C_Hose reads the hose input file 6 7
4 8 9
and builds a model of the hose in ADAMS command language. 10
The pre-processor creates point masses along the hose at each 3
input point and connects the masses with beam elements. The 2
beam elements contain the stiffness properties of the hose and 1
the mass of the hose. The mass for each beam element is
calculated by dividing the total hose mass by the number of Figure 4. Placing of markers on example hose sections.
beam elements. If there is a change in length specified, the
starting length of the beam elements are defined to be longer or After the markers are defined for each hose section,
shorter than the actual distance between the point masses. This C_Hose specifies a rotation for one of the markers at each
will cause the hose to expand or shrink the desired amount interior hose point. Interior hose points include all hose points
when the ADAMS Solver simulation is executed. If there is a except for the two end points. This rotation aligns the markers
clamp defined at any of the input points, the position of the at the ends of the hose sections so that tangent continuity (C1 )
point mass at that point is constrained. This ADAMS model of exists between sections. An example of a rotation can be seen
the hose is written to a file that is read into ADAMS Solver in Fig. 4. In this example, marker number 6 is rotated to align
during the simulation step. with marker 5. This rotation is specified in the ADAMS input
file but does not actually take place until the ADAMS Solver
In order to approximate the initial path of the hose, the pre- simulation is run. After this initial rotation aligns the markers,
processor assumes that adjacent hose points in the input file are no more relative rotation is allowed between those two markers
connected by straight sections of hose called hose sections. for the rest of the simulation, creating a C1 continuous hose.
Figure 3 gives an illustration of the straight hose sections These marker rotations are necessary so that the hose model
between each hose point. The hose points at the start and end will behave like a physical hose. Creating the hose in this
of the entire initial hose path are called end points. Interior manner causes it to return to straight if all clamping conditions,
hose points include the set of all hose points except for the two hose end forces, and gravitational forces are removed.
end points.
The next step is to define beam elements between adjacent
markers on the same hose section. For example, in Fig. 4,
beam elements connect markers 1 and 2, 2 and 3, 3 and 4, 4 and
5, 6 and 7, 7 and 8, 8 and 9, and 9 and 10. The beam element
transmits forces between markers according to beam theory. It
is within the beam element definition that the bending, axial,
and torsional stiffness is entered. An arbitrary damping ratio of
0.1 is also included in each beam element in order to prevent
the hose from oscillating during ADAMS simulation.
Figure 3. An initial hose path divided into two hose sections.
An additional feature of C_Hose is the ability to change the
The next step after reading in the input information is to length of a hose by a percentage of its original length. This
define ADAMS markers along each of the straight hose feature uses the beam element definition step to change the
sections. A marker is a local coordinate system that is defined length of a hose in the following way. When a beam element is
by its origin position and orientation with respect to the global defined, its length must be specified along with the two end
markers and the stiffness. If the specified length is equal to the

3 Copyright 2002 by ASME

actual distance between the two markers it connects, the beam ADAMS outputs the position of each hose point to a file. The
element length is in equilibrium when the ADAMS simulation hose-modeling program post-processor then reads this file,
starts and the total axial length of the hose will remain the finds the final positions of each hose point, and writes these
same. The length of the hose is changed by specifying a beam positions to an output file. This output, in the form of the
length that is longer or shorter than the distance between the (x,y,z) positions of the hose points in their equilibrium
end markers. When the ADAMS simulation is run, the positions, is an approximation of the true path of the hose.
elements are not in equilibrium and they will change length
until they reach equilibrium, therefore changing the length of VERIFICATION OF HOSE MODEL
the hose to the desired length. The hose-modeling program was verified for two different
configurations of an example hose. The hose paths generated
After the beam ele ments are defined, C_Hose places point by ADAMS were compared with a real hose for both
masses at the origin of each marker along the hose, shown as configurations. These tests verified the ADAMS bending
black circles in Fig. 5. The mass of each hose section (m) is model. The torsional and axial models were not verified
calculated by multiplying the length of the section (L) by the because they are not as imp ortant as bending in hose
mass per unit length for the hose. The section mass (m) is then applications. For both models, the paths ADAMS predicted
divided up between the point masses for the section as shown in were very similar to the actual paths for the real hoses.
Fig. 5. The mass is divided so that there is a uniform mass
distribution along the hose. The first test configuration was a cantilevered hose. One
end of a real hose was clamped into a vise with the hose held
parallel to the ground. The hose end was then released and the
deflection of the free end was measured. To verify the
ADAMS hose model, the hose was entered into C_Hose and
the deflection of the free end due to gravity was noted. Finally,
L analytical expressions for the deflection of a cantilever beam
m/8 were used as an additional verification method. Table 1 shows
that the deflections compared favorably.
m/4
Table 1. Results of the cantilevered hose verification.
m/4
m/4 C_Hose Cantilever
Real hose
model Beam Formula
m/8
Deflection
Figure 5. Distribution of mass for a hose section. of the free Approx. 50 51.3 51.75
end (mm)
Mass moments of inertia for the hose sections are also
calculated and assigned to the point masses. These moments of
inertia are required for an ADAMS dynamic simulation. Two The second test configuration was a hose initially
different mass moments of inertia are calculated for each hose positioned on a horizontal table (no gravity effects deflections).
section: 1. Mass moment of inertia for a hose with length L/j, Using an x-y coordinate system, the hose is placed along the
mass m/j, and relative to a coordinate system at the center of the line y = x. One end of the hose was fixed at the origin and the
hose; and 2. Mass moment of inertia for a hose with length center-point of the other end was moved in the negative y
L/2j, mass m/2j, and relative to a coordinate system at the end direction. The moving end of the hose was free to rotate as it
of the hose. The variable j stands for the number of beam moved. The deformed hose path calculated by ADAMS was
elements per hose section. Mass moment of inertia #1 is plotted at full scale and the real hose was laid on top of the
assigned to the interior point mass locations on the hose section ADAMS hose. Figure 6 shows two pictures of the hose laid
and mass moment of inertia #2 is assigned to the two end point onto the drawing of the predicted path. Figure 6(a) shows the
mass locations for the hose section. undeformed hose and Fig. 6(b) shows the same hose after it had
been deformed. The real hose shown in Fig. 6 includes metal
The last step in the pre-processor is fixing the ends of the end fittings, which were not part of the model. The model only
hose and fixing the hose at any clamp points that were includes the black part of the real hose. The real hose took
specified. Both markers at the end points of the hose and any almost exactly the same path as the ADAMS model. Note that
markers associated with specified clamps are locked to the the movement of the right end of this hose required some
global reference frame. This constraint does not allow any additional modeling in ADAMS since C_Hose doesnt have the
movement or rotation of the marker during the ADAMS capability to model moving ends. C_Hose was used to create
simulation. the model of the straight hose and that model was loaded into
ADAMS View. The motion of the end of the hose was
ADAMS SIMULATION AND C_HOSE POST- prescribed within ADAMS View and the resulting hose shape
PROCESSOR was plotted from ADAMS View. This test verified that the
After the C_Hose pre-processor has created the ADAMS ADAMS beam elements, defined by the C_Hose preprocessor,
Solver model of the hose, an ADAMS Solver static equilibrium modeled the hose path correctly even though the C_Hose pre-
simulation is performed. After the simulation is executed, processor could not create the motion required by the test.

4 Copyright 2002 by ASME

 can be used by hose designers to predict the real paths of hoses
before the first prototype build, allowing them to identify and
correct interference problems earlier in the design process. The
hose model could be improved by adding the ability to model
hoses with moving end points.

ACKNOWLEDGEMENTS
The authors would like to thank Deere & Company and the
Virtual Reality Applications Center for supporting this work.

REFERENCES
[1] Mechanical Dynamics Incorporated, 2001, ADAMS
Users Manual V10.

[2] Elliott, A. S., 1996, "Hydraulic Hose Validation

Report, Internal Document, Mechanical Dynamics
Figure 6. Real hose compared to plot of Corporation.
ADAMS hose model.
CONCLUSION [3] Rodriguez, J. and Keil, M., 2001, CAD Methodology
This paper describes a hose-modeling program that was for design of hoses and other flexible elements,
shown to accurately model two example hoses. C_Hose Proc., 6 th Annual International Conference on
requires the input of an initial hose path, clamping constraints, Industrial Engineering Theory, Applications, and
and hose properties. The C_Hose pre-processor uses this Practice, San Francisco, CA, November 18-20.
information to construct an ADAMS model, which is then
analyzed. Finally, the C_hose post-processor creates a results [4] Chipperfield, K. and Vance, J., 2002, Stiffness Testing
file that gives the approximate path for the hose. This program of Hydraulic Hoses, Proc., ASME: 28th Design
Automation Conference, in press.

5 Copyright 2002 by ASME