// 5-MINUTE PAPER TALK
Functional Prototype of a Winding Machine for Manufacturing Hollow Tubes with Synthetic or Natural Fibers
David Sandoval Núñez · Juan Pablo Davila Moncayo
Álvaro José Rojas Arciniegas, Ph.D.
Universidad Autónoma de Occidente · Cali, Colombia
// 01 MOTIVATION
Why Study Fiber-Wound Composite Tubes?
2
Demand for lightweight structures
Automotive and aerospace sectors seek lightweight, high-performance components. Fiber orientation is a key design variable.
Natural fibers: research potential
Renewable origin, low density, biodegradability and recycling potential are discussed in the paper.
Indicative material densities (g/cm³)
Natural fibers
1.3–1.5
Aluminum
≈ 2.7
Steel
≈ 7.8
Paper Fig. 1 · Circumferential, helical and polar winding
The manufacturing challenge
Commercial winding systems are costly and complex, restricting laboratory access.
// 02 RESEARCH GAP
Prior Systems and COILER’s Contribution
3
Study
Axes
Fiber
Trajectory
Control
Mateen et al. [4]
2
Glass
Helical
AVR
Quanjin et al. [6]
3
Carbon
Helical
Arduino UNO
Andrianov et al. [5]
4
Carbon
Geodesic
G-code
Kivirenko et al. [3]
2
Glass
Helical / circum.
CNC
COILER (this paper)
3
Glass + hemp
Helical + indexing
Pi 4 + Arduino Mega
Research gap
Simplified control + fiber-width-based indexing + validation with both glass and hemp.
// 03 METHOD
Three Degrees of Freedom and Mechanical Design
4
Paper Fig. 2 · CAD model and motion directions
Paper Fig. 5 · Constructed prototype
Paper Fig. 3 · Fiber guide system
M1
Mandrel rotation
M2
Longitudinal carriage
M3
Fiber-guide orientation
// 04 SYSTEM IMPLEMENTATION
Electronic Architecture and Motion Control
5
Paper Fig. 4 · Control, power, actuation and sensing architecture
PROCESSING
Raspberry Pi 4: user interface and process parameters
MOTION CONTROL
Arduino Mega 2560: STEP/DIR to NEMA 23 and NEMA 17 motors
FEEDBACK & SAFETY
AS5600 magnetic encoder + carriage endstop sensors
Drivers: DM542 and TB6560
// 05 MODEL
From Parameters to Helical Trajectories
6
Helical pitch · Eq. (1)
P = πD / tan(α)
Mandrel speed · Eq. (3)
RPM(M1) = 60 V(M2) tan(α) / πD
Indexing step · Eq. (9)
Δθ = 360° / N
Paper Fig. 6 · Indexed trajectories: glass (A) and hemp (B)
Why indexing matters
Glass: 7 mm tape
Hemp: 1 mm thread
Narrower fiber → finer angular spacing and more positions per layer.
// 06 VALIDATION
Motion Validation: Speed and Angular Accuracy
7
Synchronized speed at 60°
Paper Table II · 500-mm useful length; transition time = 4
Mandrel
M1 (rpm)
M2 (mm/s)
32.5 mm
23.98
23.56
22.2 mm
35.11
23.56
Smaller mandrel → higher M1 speed
Measured angular positioning
Paper Table III · Encoder readings, all targets = 60°
Fiber
Dm (mm)
Measured
Error
Hemp
22.2
59.52°
0.48°
Glass
22.2
59.47°
0.53°
Hemp
32.5
59.54°
0.46°
Glass
32.5
59.32°
0.68°
MAX ERROR 0.68° | 1.13%
// 07 MANUFACTURING
Manufactured Composite Tubes: Glass and Hemp
8
Paper Fig. 7 · Glass tube (top) and hemp tube (bottom)
Experimental specimens
Glass fiber
Ø 27.075 mm · L 470 mm
Hemp fiber
Ø 28.025 mm · L 320 mm
60° helical winding
2 layers · 22.2-mm initial mandrel
// 08 RESULTS
Four-Point Bending: Results and Implications
9
Paper Fig. 8 · Load–flexural strain under adapted four-point bending
Flexural comparison
Glass
Hemp
Max. force
345.95 N
109.49 N
Flexural stress
4.5 MPa
1.3 MPa
Young’s modulus
330 MPa
49 MPa
Tests reached 5% strain; tubes deformed without fracture and recovered their shape.
Validated: 3-DOF helical winding, angular accuracy and manufactured tubes.
Future: stiffer matrices + longer resin working times
Universidad Autónoma de Occidente · Cali, Colombia